Acid–Base Homeostasis in Dialysis

Hemodialysis

Hemodialysis bath solutions around the world contain widely varying bath [HCO ₃ ^_ ] values, from as low as 25 mEq/L in Japan to as high as 40 mEq/L in some patients in the United States. This variability reflects the lack of studies defining the optimal bath concentration for achieving a normal predialysis blood [HCO ₃ ^_ ] and pH. The wide range of bath values tolerated is likely due to parallel variations in buffering and/or organic acid production, a homeostatic response that minimizes the change in pH that occurs during the dialysis session. Most commonly, bath solutions contain [HCO ₃ ^_ ] values of 32–35 mEq/L.

In addition to HCO ₃ ^_ , hemodialysis bath solutions contain an organic anion, either acetate or citrate, in low concentration. These anions are produced by the reaction of their respective organic acids with HCO ₃ ^_ during the generation of the final bath solution, a chemical event necessary to keep bath pH in a range that prevents precipitation of calcium and magnesium salts. Generation of HCO ₃ ^_ from these organic anions is determined by their rate of influx and metabolism. In contrast to HCO ₃ ^_ , the transmembrane gradient for these organic anions remains constant during dialysis once influx rate equals metabolic rate ( Fig 33.1 B), and therefore, these anions are the source for as much of 35%–50% of the total HCO ₃ ^_ added during treatment, despite their low bath concentration.

In patients receiving intermittent hemodialysis, predialysis serum [total CO ₂ ], the parameter usually measured, is determined by the dialysis prescription and patient characteristics outlined in Table 33.1 . Of the patient characteristics, the first two, endogenous acid production and fluid retention between treatments, are the primary determinants of predialysis serum [total CO ₂ ]. Renal generation and loss of HCO ₃ ^_ are usually trivial. Organic acid production during dialysis has a notable impact on the net addition of alkali during dialysis but is a stable feature of the bath [HCO ₃ ^_ ] prescribed, and therefore, it has little effect independent of the dialysis prescription unless it is unusually large (see discussion of metabolic acidosis later in the chapter). The estimated effects of variations in endogenous acid production and fluid retention between treatments are shown in Table 33.2 .

As shown in the table, variations in endogenous acid production over a reasonable range can change predialysis serum [total CO ₂ ] from normal to frankly acidotic values. Endogenous acid production varies directly with dietary intake of sulfur-containing proteins. Thus, patients with limited intake of animal protein and grains will have a higher predialysis serum [total CO ₂ ] than patients eating a diet high in these foods. Variations in fluid retention have a smaller but clinically significant effect.

Hemodialysis three times weekly does not restore predialysis serum [total CO ₂ ] to 24 mEq/L unless endogenous acid production is low. Before 2005, the average predialysis value ranged from 19 to 23 mEq/L in stable outpatients, but more recently, the range has risen to 23–26 mEq/L, suggesting that endogenous acid production has fallen. The cause is unknown but seems likely to be due to a difference in diet, reflecting an older and more debilitated hemodialysis population. It is important to emphasize that predialysis [total CO ₂ ] is a nadir value. Serum [total CO ₂ ] rises to 28–32 mEq/L just after each hemodialysis treatment and gradually falls to the predalysis level over the interval between treatments. Not surprisingly, the lowest predialysis value occurs after the longest interval (68 hours) between treatments. At the end of the shorter interval (44 hours), the value is approximately 1 mEq/L higher.

Two large cohort studies have demonstrated an increased mortality risk both in patients with low (< 19 mEq/L) and high (> 26 mEq/L) predialysis serum [total CO ₂ ] values. In the former group, the increased risk appears to be directly attributable to the presence of metabolic acidosis. In the latter group, the increased risk is associated with alkalosis but appears to be primarily caused by negative nutritional and inflammatory factors. Predialysis serum [total CO ₂ ] can be varied by adjusting bath [HCO ₃ ^_ ], and short-term studies have shown that in patients with low values (< 20 mEq/L), raising the value sufficiently to increase serum [total CO ₂ ] to 24 mEq/L has beneficial effects on both muscle and bone metabolism. Although this practice has been widely adopted in the United States, a more recent large cohort study suggests that increasing bath [HCO ₃ ^_ ] is an independent risk factor for death, raising the likelihood by 8% for every 4 mEq/L increase. Increasing bath [HCO ₃ ^_ ] appears to increase the buffering and metabolic response to rapid alkali addition during treatment, and the latter is a potentially maladaptive response.

Peritoneal Dialysis

Peritoneal dialysis is a continuous therapy and is therefore much less disruptive from an acid-base perspective. With this form of renal replacement therapy, steady-state serum [total CO ₂ ] is also determined by the dialysis and patient characteristics outlined in Table 33.1 . The dynamic features of alkali addition during dialysis are illustrated schematically in Fig. 33.2 . Panel A in the figure highlights the fact that lactate diffuses from the bath into the patient and HCO ₃ ^_ diffuses from the patient to the bath during dialysis. Panel B shows that the time course of the transmembrane concentration gradients differs notably from hemodialysis. The gradient for lactate influx basically remains constant throughout the bath dwell period. A large gradient is maintained by rapid metabolism of the added lactate ions, generating new HCO ₃ ^_ . Blood [lactate] does not rise notably during dialysis. Immediately after instilling a fresh bath, there is a rapid efflux of HCO ₃ ^_ from patient to the bath, as it contains no HCO ₃ ^_ . The magnitude of this initial loss is inversely related to the extracellular fluid (ECF) [HCO ₃ ^_ ]. Within 15 minutes, bath [HCO ₃ ^_ ] rises to ~ 80% of the ECF concentration, and efflux continues but at a much slower rate. At the end of the dwell, net influx of lactate exceeds HCO ₃ ^_ efflux, so that ~ 50 mEq of HCO ₃ ^_ is added to the patient each day with four 6-hour dwell periods.

Bath [lactate] has empirically been set to maintain a serum [total CO ₂ ] in the normal range in most patients, at 40 mEq/L. In some centers in Europe, peritoneal dialysis with an HCO ₃ ^_ -containing bath solution (mixed just before instillation) has improved acid-base status further, but this technique is technically cumbersome and has not gained wide acceptance. The higher exchange volumes and shorter dwell times now routinely used in continuous cycling peritoneal dialysis do not change serum [total CO ₂ ] notably.

Normal Values

The ranges of “normal” values for predialysis serum [total CO ₂ ] and for blood pH, P co ₂ , and [HCO ₃ ^_ ] in patients receiving conventional hemodialysis, and the steady-state values for these parameters in patients receiving peritoneal dialysis are shown in Table 33.3 . In patients receiving hemodialysis, average serum [total CO ₂ ] has risen by 1–2 mEq/L in the last decade without any change in dialysis prescription, possibly reflecting a lower protein intake in the older population now receiving this treatment. Blood P co ₂ and pH measurements indicate either an essentially normal acid-base status or a mild acidosis with an appropriate secondary respiratory response. Patients receiving peritoneal dialysis have average values for venous serum [total CO ₂ ] in the normal range, although limited arterial acid-base measurements in these patients also show a very mild acidosis.

Diagnosis and Management of Acid-Base Disorders

The nephrologist caring for patients receiving dialysis therapy has two tasks with regard to their acid-base status. The first is to identify patients with either persistently low values for serum [total CO ₂ ] (less than 20 mEq/L) or persistently high values for serum [total CO ₂ ] (greater than 25 mEq/L) and to determine whether treatment modification can improve this situation. The second task is to uncover superimposed acute acid-base disturbances. In making either assessment, serum samples for total CO ₂ must be processed and run in a timely fashion. Shipment of blood samples to a distant laboratory and delay in analysis may spuriously reduce the value by as much as 5 mEq/L. If a low or high serum [total CO ₂ ] is validated, the next step is to obtain an arterial blood sample for measurement of pH and P co ₂ in order to fully characterize the acid-base disorder. In patients with functioning fistulas or grafts, such a sample can simply be obtained from the arterial side of the cannulated access.