Nephrol Dial Transplant (1998) 13 [Suppl 6 ]: 6265

Nephrology Dialysis Transplantation

Behaviour of acidbase control with dierent dialysis schedules

M. Feriani
Department of Nephrology, St Bortolo Hospital, Vicenza, Italy

Introduction
In a functionally anephric patient, the dialytic procedure rather than the kidneys is responsible for acid excretion and bicarbonate regeneration. Metabolic acid production due to protein and phospholipid metabolism is oset by buers provided during dialysis. Metabolism and the consequent acid production is a continuous function, while dierent dialysis techniques involve dierent time schedules. During continuous therapies [continuous ambulatory peritoneal dialysis (CAPD) and continuous arteriovenous haemoltration (CAVH )], the continuous buer ux towards the patient continuously osets the metabolic acid production and, consequently, stability of the patients acidbase status is normally recorded. In standard thrice-weekly intermittent therapies [haemodialysis (HD), haemoltration ( HF ) and haemodialtration (HDF )], the task of counterbalancing a 48 h metabolic acid production during the interdialytic period should be accomplished during the few hours of dialysis. Consequently, the acidbase status uctuates from a post-dialytic alkalosis to a predialytic acidosis. Between these two periods and for a variable length of time, a normal acidbase status is generally achieved. In daily treatments [nightly intermittent peritoneal dialysis (NIPD)], since a 24 h metabolic acid production should be oset by a 810 h dialysis, less uctuation could occur.

and 10% of patients had a metabolic alkalosis. Only ~25% of patients had a normal acidbase status (arterial bicarbonate normal range 23.526.1 mmol/l ). In the majority of studies [2,3] in which CAPD solutions containing 40 mmol/l of lactate substituted for standard 35 mmol/l lactate solutions, the correction of uraemic acidosis markedly improved ( Table 1). However, a substantial number of patients still remained acidotic, while an increased number of patients developed a metabolic alkalosis. The patients plasma bicarbonate level in CAPD depends on metabolic acid production and ultraltration [4]. Since these two parameters are highly variable in a CAPD population, a large interpatient variability is usually recorded in the CAPD population.

Intermittent treatments
In intermittent treatments, the pre-dialysis, postdialysis and interdialysis acidbase status should be taken into account.

Pre-dialysis acidbase status

Table 2 reports pre-dialysis blood bicarbonate of patients in which bicarbonate HD was performed. A large interpatient and interstudy variability can be seen [511]. The pre-dialysis bicarbonate is modulated by three variables: metabolic acid production, ultraltration and the amount of buer infusion dependent on the dialytic procedure. When the dialysis procedure (blood ow, dialysate ow, bicarbonate bath concentration, lter) and ultraltration are constant, a balance between metabolic
Table 1. Acidbase changes with CAPD solutions of dierent lactate content Parameter Venous TCO 2 Venous TCO 2 Buer Lactate Lactate 35 mmol/l 23.402.80 19.253.15 40 mmol/l 27.402.80 26.003.29 Reference 2 3

Continuous treatments
CAPD treatment usually maintains a stable acidbase status. In a recent multi-centre cross-sectional analysis of 75 stable CAPD patients treated with conventional 35 mmol/l lactate solution [1], the median value of arterial plasma bicarbonate was 22.7 mmol/l and quartiles were 19.9 and 24.7 mmol/l. About 60% of patients had varying degrees of metabolic acidosis; 25% of patients had a bicarbonate level <19.9 mmol/l
Correspondence and oprint requests to: Mariano Feriani, Department of Nephrology, St Bortolo Hospital, 36100 Vicenza, Italy

Venous TCO normal range 28.51.8. 2

1998 European Renal AssociationEuropean Dialysis and Transplant Association

Acidbase control with dierent dialysis schedules Table 2. Pre- and post-dialysis blood bicarbonate in dierent studies Pre-dialysis Post-dialysis Dialysis bicarbonate concentration 36 mmol/l 35 mmol/l 33 mmol/l 36 mmol/l 30 mmol/l variable 32 mmol/l 36 mmol/l No. of patients Reference

63 Table 3. Theoretical acidbase balance in bicarbonate haemodialysis Metabolic acid production (mmol/week) Organic anion loss during dialysis Weekly base required (mmol/week) Dialytic base gain required (mmol/ dialysis) three dialyses/week Blood bicarbonate (mmol/l ) Bicarbonate gain (mmol/session) Base gain 240 (mmol/session) 420 300 720 240 16 360 120 18 20 315 270 75 30 22 24 225 180 15 60

19.80.29 19.10.85 18.43.0 22.02.4 21.52.0 21.04.0 17.4 18.0

24.20.3 24.70.52 27.34.0 26.03.0 25.7 26.7

16 10 34 8 5 5 9 9

5 6 7 8 9 10 11 11

Dialysis duration, 4 h; bicarbonate concentration in the bath, 32 mmol/l; bicarbonate dialysance, 150 ml/min. From reference [13]. Table 4. Inuence of dialysis schedule on pre-dialysis blood bicarbonate Blood bicarbonate mmol/la Bicarbonate HD Lactate HDF Bicarbonate HDF aReference [16 ]; breference [17]. 15.82.7 18.63.3 21.02.1 Buer ux mmol/minb

1.05 1.56 2.06

Fig. 1. Relationship between pre-dialysis blood total CO and daily 2 acid production.

acid production and dialytic bicarbonate gain is achieved, and a stable pre-dialysis blood bicarbonate should be recorded [12]. The level of pre-dialysis blood bicarbonate is determined by the amount of acid production (metabolic acid production and base lost across the dialytic membrane) (Figure 1). Since dialytic bicarbonate gain is correlated inversely with the blood bicarbonate, if the patient for any reason has a low pre-dialysis blood bicarbonate, dialytic bicarbonate gain increases and overcomes the acid production. Consequently, since the balance is positive, the blood bicarbonate should be higher at the next dialysis. This reduces dialytic base gain until dialytic base gain equals acid production [13]. This feedback mechanism also occurs when the pre-dialysis blood bicarbonate concentration is greater than that achieved at the equilibrium point. An example of this mechanism is reported in Table 3. This feedback implies that in a stable patient the hydrogen ion balance between two dialysis sessions is neutral [13]. Ultraltration reduces pre-dialysis blood bicarbonate by reducing dialytic base gain. It could be calculated theoretically that 1 l of uid removal reduces the dialytic base gain by ~20 mmol/session [14]. Although these convective losses do not change the bicarbonate concentration during dialysis, the subsequent retention of an equivalent amount of alkali-free uid in the interdialytic period does. This theoretical phenomenon is demonstrated in clinical practice. In a cross-over study in 29 patients,

Fabris et al. [15] found that reducing the interdialytic weight gain of 1 kg, there was a signicant pre-dialysis blood bicarbonate increase of 1.6 mmol/l. The bicarbonate concentration in the dialysate bath greatly inuences the dialytic buer gain, the latter being a function of the concentration gradient between blood and dialysate bicarbonate concentration. Seyart et al. [11] demonstrated that by increasing the bicarbonate content in the dialysate from 32 to 36 mmol/l, a signicant increase in pre-dialysis blood bicarbonate could be achieved. This aspect is very important in mixed diusive convective treatments since, in these treatments, the concentration of buer and the administration rate of the replacement uid could easily be changed. Table 4 reports the pre-dialysis blood bicarbonate changes in 11 patients treated consecutively for 3 months with three dierent dialysis schedules [16 ]. These changes were correlated with the buer ux calculated for each dialytic schedule for the same blood bicarbonate [17]. Post-dialysis acidbase status The post-dialysis blood bicarbonate depends on the distribution space for bicarbonate that is a function of both pre-dialysis blood bicarbonate and body weight, and on the bicarbonate concentration in the dialysate or the amount of buer infused with the replacement uid [18]. During the session, bicarbonate is transferred towards the patient following the concentration gradient between blood and dialysate. The gained buer distributes into the bicarbonate space and increases the blood bicarbonate concentration. This increase in

64

M. Feriani

blood bicarbonate decreases the transfer of bicarbonate from dialysate by decreasing the concentration gradient across the membrane ( Figure 2). When blood bicarbonate concentration tends to equal dialysate bicarbonate concentration, no further bicarbonate ux from the dialysate occurs. However, in a standard HD, >8 h are required to achieve this equilibration point. In Figure 3, the theoretical behaviour of blood bicarbonate during a standard HD session is depicted. Two dierent distribution spaces for bicarbonate are compared. A linear increase in blood bicarbonate is recorded for the large distribution space throughout the whole dialysis session, while, for the small distribution space, the line tends to reach a plateau during the session. In patients with a large distribution space, the dialytic bicarbonate ux during the session is diluted into a large distribution space, while the dialytic bicarbonate ux rapidly saturates the small distribution space in low body weight patients. These latter patients usually experience a post-dialysis alkalosis. The large interpatient post-dialysis variability is also conrmed in a clinical setting ( Table 2).

Interdialytic acidbase status During the interdialytic period, blood bicarbonate decreases according to the metabolic acid production and distribution space for bicarbonate. In Figure 4, the theoretical decrease in blood bicarbonate during the interdialytic period is depicted. Dierent lines represent dierent metabolic acid productions and dierent distribution spaces for bicarbonate. There is a general conviction that, in a clinical setting, blood bicarbonate concentration decreases linearly at the rate of ~2 mmol per day. In Figure 5, the time course of two patients treated with a standard bicarbonate HD (4 h, dialysate bicarbonate 35 mmol/l, dialysate acetate 5 mmol/l ) is reported. It can be seen that blood bicarbonate increased after the dialysis session, probably because of the acetate metabolism, and then a sharp decrease in blood bicarbonate concentration was registered. This eect was probably due to the dilution since, in the rst hours after dialysis, patients usually experience maximum thirst. During the rst interdialytic day, the rate of decrease in blood bicarbonate concentration was linear and corre-

Fig. 2. Theoretical bicarbonate ux (&) and blood bicarbonate increase (+) during dialysis.

Fig. 4. Theoretical decrease of blood bicarbonate during the interdialytic period. &, a 40 l distribution space for bicarbonate and a 50 mmol/day acid production; +, a 20 l distribution space for bicarbonate and 50 mmol/day acid production; ,, a 40 l distribution space for bicarbonate and 150 mmol/day acid production; and 2, a 20 l distribution space for bicarbonate and 150 mmol/day acid production.

Fig. 3. Theoretical behaviour of blood bicarbonate during a standard hamodialysis (dialysate bicarbonate 35 mmol/l ). &, a 21.7 l distribution space for bicarbonate; +, a 43.5 l distribution space.

Fig. 5. Blood bicarbonate concentrationss of two patients during a dialysis session and an interdialytic period.

Acidbase control with dierent dialysis schedules

65

sponded to the theoretical calculated decrease. For the subsequent 16 h, no changes were recorded. It could be hypothesized that the intervention of intracellular and bone buers contributed to maintaining a constant blood bicarbonate. Even though only anecdotal, this acidbase prole suggests that the base balance between two dialysis sessions is not neutral as hypothesized by Gennari [13] but that endogenous body buers are used to counterbalance the blood bicarbonate decrease below physiological levels.

The frequency of treatment exerts a strong inuence on acidbase uctuations and, consequently, an increase in treatment frequency should improve the acidbase status as a whole, since high frequency treatments better approximate the physiological functions.

References
1. Feriani M. Buers: bicarbonate, lactate, pyruvate. Kidney Int 1996; 50 [Suppl 56 ]: S75S80 2. Nolph KD, Prowant B, Serkes KD et al. Multicentric evaluation of a new peritoneal dialysis solution with a high lactate and low magnesium concentration. Peritoneal Dial Bull 1983; 3: 6365 3. Mandelbaum JM, Heistand ML, Schardin KE. Six months experience with PD-2 solution. Dial Transplant 1983; 12: 259260 4. Feriani M, Ronco C, La Greca G. Acid base balance with dierent CAPD solutions. Peritoneal Dial Int 1996; 16 [Suppl 1]: S126S129 5. Hakim RM, Pontzer MA, Tilton D, Lazarus JM, Gottlieb MN. Eects of acetate and bicarbonate dialysate in stable chronic dialysis patients. Kidney Int 1985; 28: 535540 6. Henrich WL, Woodard TD, Meyer BD, Chappel TR, Rubin LJ. High sodium bicarbonate and acetate hemodialysis: doubleblind crossover comparison of hemodynamic and ventilatory eects. Kidney Int 1983; 24: 240245 7. La Greca G, Feriani M, Bragantini L, Petrosino L, Santoro A, Altieri P. Eects of acetate and bicarbonate dialysate on vascular stability: a prospective multicenter study. Int J Artif Org 1987; 10: 157162 8. Bijaphala S, Bell AJ, Bennett CA, Evans SM, Dawborn JK. Comparison of high and low sodium bicarbonate and acetate in stable chronic hemodialysis patients. Clin Nephrol 1985; 23: 179183 9. Malberti F, Surian M, Colussi G, Minetti L. The inuence of dialysis uid composition on dialysis tolerance. Nephrol Dial Transplant 1987; 2: 9398 10. Heinechen FG, Brady-Smith M, Haynie J, Van Stone JC. Prescribing dialysate bicarbonate concentration for hemodialysis patients. Int J Artif Org 1988; 11: 4550 11. Seyart G, Ensminger A, Scholz R. Increase of body mass during long-term bicarbonate hemodialysis. Kidney Int 1987; 32 [Suppl 22]: S174S177 12. Gennari FJ, Rimmer JM. Acid base disorders in end-stage renal disease: part I. Semin Dial 1990; 3: 8185 13. Gennari FJ. Acid base balance in dialysis patients. Kidney Int 1985; 28: 678688 14. Sargent JA, Gotch FA. Bicarbonate and carbon dioxide transport during hemodialysis. J Am Soc Artif Intern Org 1979; 2: 6172 15. Fabris A, La Greca G, Chiaramonte S et al. The importance of ultraltration on acid base status in a dialysis population. J Am Soc Artif Intern Org 1988; 34: 200201 16. Biasioli S, Feriani M, Chiaramonte S et al. Dierent buers for hemodialtration: a controlled study. Int J Artif Org 1989; 12: 2530 17. Feriani M, Ronco C, Biasioli S, Bragantini L, La Greca G. Eect of dialysate and substitution uid buer on buer ux in hemodialltration. Kidney Int 1990; 39: 711717 18. Santoro A, Spongano M, Ferrari G et al. Analysis of the factors inuencing bicarbonate balance during acetate free bioltration. Kidney Int 1993; 43 [Suppl 41]: S184S187 19. Bergstrom J, Alvestrand A, Furst P. Plasma and muscle free amino acids in maintenance hemodialysis patients without protein malnutrition. Kidney Int 1990; 38: 108114 20. Lefebvre A, de Vernejoul MC, Gueris J, Goldfarb B, Graulet AM, Morieux C. Optimal correction of acidosis changes progression of dialysis osteodystrophy. Kidney Int 1989; 36: 11121118 21. Bazilinsky NG, Dunea G, Ing TS. Treatment of metabolic alkalosis in renal failure. Int J Artif Org 1987; 10: 284286

Clinical relevance of transient acidbase disturbances

Several clinical studies have demonstrated that acidosis, even when mild and transient, aects organs and physiological functions. Bergstrom et al. [19] have demonstrated that the muscle valine concentration in dialysis patients is correlated with the pre-dialysis blood bicarbonate concentration. Only patients with normal acidbase status had normal values of this amino acid. Changes in bone formation parameters during 18 months were studied in two groups of patients, one group with reduced and one with increased bone formation rate [20]. Correction of pre-dialysis acidosis led to an increase in plasma osteocalcin and bone formation rate in the low bone formation rate group, while, in the high bone formation rate group, the progressive increase of osteocalcin was retarded. In addition, the correction of acidosis reduced plasma parathyroid hormone. There are very few data in the literature about the clinical consequences of alkalosis and no data about the clinical consequences of chronic intermittent alkalosis in dialysis patients. Severe alkalosis depresses the central nervous system and increases neuromuscular excitability [21]. It causes hypokalaemia and cardiac arrhythmias and enhances digitalis intoxication. Alkalosis also increases binding of oxygen to haemoglobin, preventing the release of oxygen to peripheral tissues. These eects were described in acute alkalosis; however, post-dialysis alkalosis is also an acute alkalosis.

Conclusion
The acidbase status of dialysis patients is determined by the dialytic schedule. The most physiological correction is achieved by continuous treatments because metabolic acid production is a continuous body function. Acidosis and possibly alkalosis have a clinical relevance in dialysis patients, even if they are mild and transient. In intermittent treatments, certain degrees of acidosis and/or alkalosis are unavoidable because in a few hours of treatment dialytic base gain should provide the buer reservoir required for the long interdialytic period.