Hemodialysis with acetate, DL-lactate and bicarbonate: A hemodynamic and gasometric study

Kidney International, Vol. 46 (1994), Pp. 1167—1177

Hemodialysis with acetate, DL-lactate and bicarbonate: A hemodynamic and gasometric study JOSE A. HERRERO, JuAN I. TROBO, JAIME TORRENTE, ANTONIO TORRALBO, FERNANDO TORNERO, ANTONIO CRUCEYRA, FRANcIsco CORONEL, and ALBERTO BARRIENTOS Nephrology and Experimental Medicine and Surgeiy Services, Hospital Universitario San Carlos, Madrid, Spain

Hemodialysis with acetate, DL-lactate and bicarbonate: A hemodynamic and gasometric study. Using invasive techniques we have studied various hemodynamic and gasometric parameters in the course of hemodialysis (HD) with different buffers in an animal model. HD sessions of 180 minutes at zero ultrafiltration were carried out on three groups of

others, some of the main mechanisms for the hypoxemia induced by this buffer reported in the literature. Bicarbonate would be the ideal buffer in HD, firstly because it is the physiological buffer of the intercellular space and secondly eight uremic dogs each, under anesthesia and constant mechanical because it would not have the side effects attributed to acetate [2, ventilation. The three groups differed only in the buffer used: acetate 19]. This has led to HD with bicarbonate progressively replacing (Group AC), equal proportions of DL-lactate and acetate (Group that with acetate, particularly in high-efficiency HD, so that, at AC+LA), and bicarbonate (Group BC). No hemodynamic changes were present, 62% of all patients in Europe are dialyzed with bicarbonseen in Group BC. In the AC and AC+LA groups we observed on minute ate [20]. Nevertheless, and in spite of technological advances, 1 a decrease of the mean blood pressure (MBP) and of the systemic vascular resistances (SVR). These parameters returned to baseline values bicarbonate HD shows some disadvantages when compared with within the first 30 minutes in Group AC+LA. In Group AC the SVR also acetate HD, such as its higher cost, greater technical problems returned to baseline values after the minute 30, but the MBP remained (precipitation in presence of calcium and magnesium), and storbelow baseline throughout the study period, together with cardiac index and left ventricular stroke work index decreases. Only in Group AC did we age problems due to its tendency to decompose into CO2 and water. see a flattening of the ventricular function curves. Only in this Group was there a decrease of the arterial oxygen pressure (Pa02) with an associated Other buffers, different from acetate and bicarbonate, have increase of the alveolo-arterial and arterio-venous 02 differences. The 02 been proposed for use in HD; among them, lactate has attracted consumption was not modified in any of the groups. Acetate as a single the greatest interest [21—24]. Thus, some clinical studies indicate buffer induces hemodynamic instability through peripheral vasodilation and reduction of myocardial contractility. The myocardial depression induced by acetate, in its turn, causes a reduction in Pa02. The mixed acetate-I-lactate buffer is hemodynamically better tolerated than acetate as single buffer, as it induces only vasodilation.

In 1976, Novello, Kelsch, and Easterling described a complex of hypotension, hyperacetatemia and intradialysis decrease of blood

bicarbonate in a group of chronic patients who they termed "acetate-intolerant" [1]. It is generally accepted that acetate has a peripheral vasodilating effect [2—7], but its actions on myocardial

contractility are still debated. For some authors acetate has a myocardial depressant effect [7—11], for others it has no action

whatsoever [12], and for still others a physiological cardiac response could be attributed to acetate's vasodilating effect, or

that use of lactate as a single buffer in the dialysis fluid is

associated to a lower incidence of symptoms as compared with acetate, and that it is comparable to bicarbonate even in highefficiency HD [23]. Lactate, as single buffer in HD, also has some disadvantages. High concentrations of lactate, such as those used in HD fluids, tend to precipitate in the presence of calcium and magnesium salts, particularly at cold temperatures, though to a much lower degree than bicarbonate [23]. Also, high concentrations of lactate in HD fluids may cause hyperlactatemia in some patients, particularly those with hepatic dysfunction and even more so in high-efficiency dialysis [23]. Mixtures of lactate and acetate in HD fluids have been studied, which logically reduce the amount of the former as compared to its use as single buffer and, furthermore, prevent high lactate concentrations in the HD fluids [21, 22, 24].

The aim of the present work was twofold: firstly, to study the discrepancies present in the literature as to hemodynamic behavimplicated as the cause of the hypoxemia frequently seen during ior—particularly myocardial contractility—and oxygenation pathe hemodialysis (HD) session [15—18]. Alveolar hypoventilation due either to CO2 loss in the dialysate [15, 16] or to a lower CO2 rameters in HD with acetate; secondly, to assess the hemodygeneration during acetate metabolism [17] and increased 02 namic and gasometric response to HD with an equiproportional consumption derived from acetate metabolism [18] are, among mixture of acetate and lactate. even a positive inotropic effect [3—6, 13, 14]. Acetate has also been

Methods Received for publication May 13, 1993 and in revised form May 11, 1994 Accepted for publication May 12, 1994

© 1994 by the International Society of Nephrology

Design of the experiment

HD was performed on 24 uremic dogs distributed in three groups of eight dogs each. Dialysis characteristics were the same

1167

1168

Herrero et at: Effects of buffer in hemodialysis

Table 1. Hemodynamic baseline values Acetate MBP mm Hg HR beats/mm

CI liter/min/m? SI mI/beat/rn2

LVSWI g minIm2 MPAP mm Hg CVP rum Hg PCWP mm Hg SVR dynes s/cm5 PVR dynes s/cm5 SI/PCWP LVSWI/PCWP

108.8 145 5.18 35.75 53.42 14,17 5.62 6.87

6.7 3.1 0.64 4.4 7.47 1.26 0.98 0.91

Acetate + Lactate 115.5

6.8

152 9.8 5.24 0.34

35.15 3.08 55.64 6.08 14.5 6.37 7.25

0.95

0.62 0.86

mmol/liter, glucose 1.5 g/liter, theoretical bicarbonate 39 mmol/ Bicarbonate 115.0

8.2

4.88

0.33 1.73 4.60 1.06 0.77 0.72

160 5.7

30.43 48.05 14.12 6.25

7.12

2106 265

1962 245

2211 164

151.9 5.6 8.1

128.7

140.8

28.46 0.61 0.80

15.29

5.3 0.47 7.84 0.52

4.44 6.78

18.06 0.29 0.39

liter, bicarbonate measured at the time of HD 34.52

0.44

mmol/liter, and acetate 4 mmol/liter. An HD device with volumetric ultrafiltration control was used

(Toray TR-321, Toray Medical Co. Ltd., Tokyo, Japan). The vascular access for HD was through the femoral vein and artery with pediatric lines (Gambro Dialysatoren GmbH & Co. KG, Hechingen, Germany) and with catheters inserted through one cutdown. We chose a biocompatible polyacrylonitrile membrane to prevent the gasometric changes induced by bioincompatible

cellulosic membranes (Filtral 8, Hospal Industrie, Meyzieu, France). The dialysate temperature was 37.5°C for all HDs. The blood flow was 8 ml/kg/min, and the dialysis fluid flow 500 ml/min.

Data are means SEM. None of the baseline values differed significantly

from each other. Abbreviations are: MBP, mean blood pressure; HR, heart rate; CI, cardiac index; SI, stroke index, LVSWI, left ventricular stroke work index; MPAP, mean pulmonary arterial pressure; CVP, central venous pressure; PCWP, pulmonary capillary wedge pressure; SVR, systemic vascular resistance; PVR, pulmonary vascular resistance.

in all cases, but for the buffer used: in Group AC acetate was used as the only buffer; in Group AC+LA an equiproportional mixture of acetate and DL-lactate was used; and in Group BC bicarbonate was the single buffer. All dialyses lasted for 180 minutes, with zero

Hemodynamic monitorization

For blood pressure measurements, the contralateral femoral artery was dissected and canalized with a Seldicath 3876 E catheter (Laboratoire Pharmaceutique, Saint-Leu-la-Fôret Cedex, France); a pulmonary artery was canalized through the dissected femoral vein (also contralateral to the HD access) with an Elecath 73-6067 Swan-Ganz catheter with thermistor (ElectroCatheter Corporation, Rahway, New Jersey, USA). The systemic

arterial, pulmonary arterial, and right atrial pressures were recorded in a Lifescope 6 monitor (Nihon Kohden Corporation, ultrafiltration and at identical dialysate temperatures. The ani- Tokyo, Japan). The heart rate (HR) was derived from Lead II of mals were under general anesthesia and constant mechanical the ECG. The cardiac output (CO), expressed in liter/mm, was ventilation. Hemodynamic parameters were assessed by invasive measured by thermodilution with an Elecath COC 4000 computer. Three successive measurements were carried out and the techniques. Preparation of the animals All the studies were carried out in dogs in which liver disease

had been previously ruled out (normal levels of alkaline phosphatase, total bilirubin, ALAT, ASAT and LDH). The mean weight of the animals was 24.7 2.28 kg in Group AC, 27.6 1.68 kg in Group AC+LA, and 24.1 1.39 kg in Group BC (intergroup differences non-significant). Acute renal failure was induced in all animals through bilateral ureteral ligation 48 hours

mean was accepted. Measurements were performed at timepoints 0, 1, 5, 15, 30, 45, 60, 90, 120, 150, and 180 (minutes after starting

HD). At each timepoint, mean blood pressure (MBP), mean pulmonary artery pressure (MPAP), central venous pressure (CVP), pulmonary capillary wedge pressure (PCWP) (all in mm

Hg), HR and CO were recorded; from these, and for the same timepoints, the following were calculated according to the formulae given [25]:

1. Cardiac index (CI) = CO/body surface area (1/mm/rn2). before HD, under antibiotic prophylaxis with gentamicin 3 mg/kg 2. Stroke volume (SV) CO• 1000/HR (ml/beat). and cephazolin 250 mg (i.v.). 3. Stroke index (SI) = SV/body surface area (mI/beat/rn2). Anesthesia was induced with pentothal sodium 10 mg/kg Lv., 4. Left ventricular stroke work index (LVSWI) = SI MBP pancuronium bromide 0.07 mg/kg i.v., and fentanyl 1 jig/kg i.v.; an 0.0136 (g min/m2). orotracheal tube was then inserted and ventilation initiated with 5. Systemic vascular resistance (SVR) = (MBP — CVP) 79.9/CO an Engstrom ER 300 volumetric respirator (LKB Medical AB, Broma, Sweden), using a gas mixture of 40% oxygen and 60%

nitrous oxide and a tidal volume of 12 mI/kg body wt. The respiratory rate was adjusted to maintain a baseline arterial CO2 pressure (PaCO2) of ca. 35 mm Hg; after reaching that PaCO2, the respirator's parameters were kept constant during the study. Spontaneous ventilation was prevented with pancuronium bromide, 0.04 mg/kg/hr i.v., which was used upon early detection of decurarization signs through the corneal reflex. Dialysis procedures

The composition of the dialysis fluids was: for Group AC, Na 138 mmol/liter, K 1.5 mmol/liter, Ca 1.75 mmol/liter, Mg 0.75 mmol/liter, Cl 109 mmol/liter, glucose 1 g/liter, and acetate 35 mmol/liter; for Group AC+LA the composition was the same, but for acetate 17.5 mmol/liter and DL-lactate 17.5 mmol/liter as buffer; for Group BC the composition was Na 139 mmol/liter, K 1.5 mmol/liter, Ca 1.75 mmol/liter, Mg 0.5 mmol/liter, Cl 106

(dynes . s/cm5).

6. Pulmonary vascular resistance (PVR) = (MBP CO (dynes . s/cm5).

—

PCWP) . 79.9/

The SI/PCWP and LVSWI/PCWP ratios were also calculated for the same timepoints. Power curves and linear regression equa-

tions were determined from individual values of LVSWI and PCWP to obtain left ventricular function curves [26]. Laboratory methods At timepoints 0, 60, 120 and 180 the plasma levels of acetate (gas chromatography, as previously described [27]) and L-lactate (enzymatic method [28]) were determined. Acetate balance was evaluated after measuring total dialysate recovered in a calibrated tank. Predialyzer and dialysate acetate were measured by triplicate and the mean was accepted. From that data the final acetate balance was extrapolated. At timepoints 0 and 180 arterial blood samples were withdrawn for measurement of hemoglobin (Hb),

1169

Herrero et al: Effects of buffer in hemodialysis

10

A

40

B

30

o

20 —10

10

C

0

a)

0

—20

C a)

><

0 —20

—30

—30 —40

20

—40

01 515304560 90 120 150 180

01

C

25

5 15 30 45 60

90 120 150 180

D

(I)

a)

0 C

0

Cl)

—10

15

t I

—20

>

.2

— —40

5

0

—50 —60

10

01 515304560 90 120 150 180

Fig. 1. Hemodynamic changes. Symbols are: (-El-) acetate;

01 515304560

90 120 150

180

Time, minutes

Time, minutes

versus minute 0: * < ØØ5. **p < 0.001; *** < 0.001.

—5

(-•-) acetate+lactate; (-h-) bicarbonate. Asterisks denote statistically significant differences

hematocrit, osmolality, sodium, potassium, calcium and BUN. The depuration efficiency of the dialysis was assessed according to

the formula KTIV = ln (BUN mm 0/BUN mm 180). At timepoints 0, 5, 15, 60, 90, 120, 150 and 180 samples were taken for gasometsy in arterial blood (intraluminal catheter) and mixed venous blood (distal lumen of the Swan-Ganz catheter). For each gasometric measurement 1 ml blood was withdrawn into a heparinized plastic syringe which was anaerobically closed and transported on ice for immediate processing in a Ciba-Corning 278 blood gas analyzer (Ciba-Corning Diagnostics Corporation, Medfield, Massachusetts, USA). For these timepoints the following oxygenation parameters were calculated [29}: 7. PAO2 = [(703 — 47) . Fi02] — PaCO2 (mm Hg). 8. P(A-a)02 = PAO2 — Pa02 (mm Hg). 9. CcO = (ITh 1.39) + PAO2 0.0031 (vol%). 10. CaO (hF3. 1.39 Sa02) + Pa02 . 0.003 1 (vol%). 11. CvO = (Hb. 1.39 Sv02) +Pv02 0.0031 (vol%). 12. C(a-v)02 = Ca02 — Cv02 (vol%).

13. Qs/Qt = CeO2 — Ca02/Cc02 — Cv02 14. DO2 = Ca02 CO/body surface area (ml O2Imin/m2). 15. V02 = D(a-v)02 CO/body surface area (ml 02/min/m2). where PAO2 is the alveolar oxygen pressure, 703 (mm Hg) the mean barometric pressure in Madrid, 47 (mm Hg) the water vapor pressure, Fi02 the inspired oxygen fraction, PaCO2 (mm Hg) the partial arterial carbon dioxide, Pa02 the partial arterial oxygen pressure, Cc02, Ca02 and Cv02 the oxygen contents in pulmonary capillary, arterial and mixed venous blood, respectively, Sa02

and Sv02, the oxygen saturation of hemoglobin in arterial and mixed venous blood, respectively, C(a-v)O2 the arterio-venous oxygen content difference, Qs/Qt the right-to-left shunt, DO2 the tissue oxygen supply, and VO2 the oxygen consumption. Statistics

Differences between variables were analyzed by the Student's

and ANOVA tests. Pearson's correlation coefficient (r)

was

computed to test the association between pairs of variables. The

1170

Herrero et al: Effects of buffer in heinodialysis

30

A 30

20

20

10

10

0

-10

a

0

—30

-10 -20

—40

—30

—50

01

10

B

5 15 30 45 60

—40

90 120 150 180

C

90 120 150 180

01

5 15 30 45 60

01

5 15 30 45 60 90 120 150

10

0

0 0

—10

0

a.

<-20

-20

-30 -40 —50 —60

—50

01 5 15 3045 60

90 120 150 180

Time, minutes

—60

180

Time, minutes

Fig. 2. Hemodynamic changes. Abbreviations are: SI, stroke index; LVSWI, left ventricular stroke work index; PCWP, pulmonary capillary wedge pressure. Symbols are as in Figure 1.

null hypothesis was rejected when P < 0.05. All values are PCWP ratios in Group AC (Fig. 2). Figure 3 shows the course expressed as arithmetic means

SEM.

Results Hemodynamic parameters

Table 1 summarizes the baseline hemodynamic values. No

over time for CVP, PCWP, MPAP and PVR. Intergroup differences. MBP decreased significantly in Group

AC as compared to Group BC on minutes 1, 5, 15 and from minute 90 of HD onwards (P < 0.01), and as compared to Group

AC+LA from minute 90 onwards (P < 0.05). However, there

intergroup differences were observed. Figure 1 shows the course were no differences in MBP between Groups AC+LA and BC over time for MBP, CI, HR and SVR. It can be seen that in throughout the study. SVR decreased in Groups AC and AC+LA Groups AC and AC+LA there was, in the first minute of HD, a as compared to Group BC on minutes 1 and 5 of HD (P < 0.01), decrease of MBP and SVR, with an increase of CI and HR. In but differences between Group AC and Group AC+LA only Group AC+LA these parameters returned to baseline levels as existed on minute 1 (P < 0.05). The CI decreased significantly in from minute 30 of HD. In Group AC SVR also returned to Group AC as compared to Groups AC+LA and BC from minute baseline values as from minute 30, but MBP remained low up to 120 onwards (P < 0.05). The SI and LVSWI decreased in Group the end of the procedure. After the first few minutes of HD with AC as compared to Group BC as from minutes 45 (P < 0.05) and acetate as single buffer there was a decrease of CI, which was as compared to Group AC+LA as from minute 90 (P < 0.05). significant as from minute 60 and maximal (25.9 5.1%) on These intergroup differences persisted when comparing the SI! minute 150 (P < 0.01). In this group HR remained increased as PCWP and LVSWI!PCWP ratios. There were no significant compared to baseline up to minute 120. We only observed a intergroup variations in HR, nor when comparing Group AC+LA decrease of SI and LVSWI, and of the SIIPCWP and LVSWII and Group BC, in CI, SI, LVSWI, SIIPCWP or LVSWI/PCWP.

1171

Herrero eta!: Effects of buffer in hemodialysis

10

B

A

8

E

a-

>

0

6

"1

ci a-

4

2 0 01

10

5 15 30 45 60

90 120 150 180

C

*

40

01 5 15 3045 60

90 120 150 180

D

30

:>

20 6 a:

0a-

90 120 150 180

*

8 E

5 15 30 45 60

01

> a-

4

10

0

2

—10

0

—20

01

5 15 30 45 60

90 120 150 180

Time, minutes

Time, minutes

Fig. 3. Hemodynamics changes. Abbreviations are: CVP, central venous pressure; PCWP, pulmonary capillary wedge pressure; MPAP, mean pulmonary artery pressure; PVR, pulmonary vascular resistances. Symbols are as in Figure 1.

Upon analysis of the ventricular function curves we observed a significant baseline correlation between LVSWI and PCWP in all

0.43 dialyzer in Group AC was 144.9 7.9 mmol/hr (6.17 mmol/kglhr), while in Group AC+LA it was 79.5 4.8 mmol/hr

three groups (Fig. 4). It can be seen that on minute 180 this (2.99 0.13 mmol/kg/hr) (P < 0.001). An inverse correlation was correlation was lost only in the group dialyzed with acetate as observed in Group AC between acetatemia and the LVSWI/ single buffer. The slopes of the regression lines constructed on PCWP ratio already from minute 60 (r = —0.83, P < 0.05). There were no differences in the baseline plasma L-lactate minutes 0 and 180 were similar for groups AC+LA and BC, and levels between the three groups (Fig. 5). The L-lactate levels significantly different for Group AC (P < 0.05; Fig. 4). increased during HD only in Group AC+LA; this increase was Acetate an4 L-lactate values In Group AC the mean maximal plasma acetate level, which had been achieved by minute 180, was 4.67 0.67 mmol/liter, the

statistically significant as compared with Groups AC and BC (P < 0.01).

highest individual value being 8.1 mmol/liter (Fig. 5); there was an

Hematocrit and other laboratory data

inverse correlation between the weights of the animals and the plasma acetate levels (r = —0.76, P < 0.05). In Group AC+LA the highest mean maximal acetate level was 1.15 0.08 mmollliter on minute 60. The plasma acetate levels were significantly higher in Group AC as compared with Group AC+LA on minutes 60, 120 and 180 (P < 0.001). The acetate absorption through the

As shown in Table 2, there were no variations in the hematocrit or the sodium levels during HD in any of the three groups. The depuration efficiency of HD, estimated through the KT/V, was similar for all three. The decrease in osmolality and potassium and the increase in calcium showed no intergroup differences.

1172

Hen-ero et al: Effects of buffer in hemodialysis

A

B

90

90

,0.78

80

80

r=0.86 r=0.78

50 40

-50

40 (I)

0

30

(I)

r=0.24

30

20

20

10

10 0 C

1

0

2 3 4 5 6 7 8 9 10 11 12

0

1

2345678

9 10 11 12

PCWP, mm Hg

PCWP, mm Hg

90 80 r=0.85

70

r'=0.77

60

0

50' 40 U)

30 20 10 0

0 1 2 3 4 5 6 7 8 9 10 11 12 PCWP, mm Hg

Blood gases

There were no differences between the three groups in the

baseline values of HC03, PaCO2, and pH (Table 3). The increase of HCO3 was significantly greater (P < 0.01) in Group BC as compared with Groups AC and AC+LA from minute 5 of HD onwards (Fig. 6). The HCO3 increase was not significantly different between Groups AC and AC+LA, although it appeared earlier in the latter. In Group BC there was an increase in PaCO2 from minute 15 of HD onwards, while this parameter showed no variations in Group AC+LA and decreased in Group AC (Fig. 6). Although the pH increased more in Group BC, the intergroup differences did not reach statistical significance (Fig. 6). There were no significant baseline differences between the

groups in Pa02, P(A-a)02, C(av)O2, DO2, V02 or QsfQt, as

Fig. 4. Ventricular function curves. Relationship between left ventricular stroke work index (LVSWI) and pulmonary capillary wedge pressure (PCWP). The regression lines are plotted from all the individual data at minute 0 (•) and at minute 180 (0). In the acetate group (A) these varied

from y =

+ 10.18 to y =

+ 18.81 (P <

0.05), in the 11.37 to y = 5.15x + 13.12 (NS), and in the bicarbonate group (C) from y = 5.49x + 8.95 toy = 5.30x 6.28x

0.91x

acetate+lactate group (B) from y = 6.lOx +

+ 843 (NS).

AC the decrease in Pa02 was associated with a fall of Pv02, with peak decrease by minute 180 (from 62.02 2.66mm Hg at minute 0 to 45.58 2.99 mm Hg at minute 180, P < 0.001). The increase

in P(A-a)02 in Group AC became significant (P < 0.05) as compared to the other two groups on minute 150, and that of C(a-v)02 on minute 120 (P < 0.05). The DO2 decrease in Group AC became significant as compared to the other two groups from

minute 120 (P < 0.05). The V02 did not show significant variations in any of the three groups, and the Qs/Qt decreased in all three (Fig. 7). Discussion

Our data show that in HD with acetate as the only buffer,

shown in Table 3; the evolution of these parameters is illustrated marked hemodynamic changes arose, consisting of initial periphin Figure 7. The Pa02 decreased only in Group AC from minute eral vasodilation followed by a decrease in cardiac output, with a 60 of HD, with a mean maximal decrease of —9.61 2.15% on blood pressure decrease over the complete dialysis session. The minute 180 (P < 0.01). This decrease became significant as vasodilator effect of acetate is well-known [2—7], yet very few compared to Group AC+LA from minute 90 (P < 0.05), and as authors report the onset of this effect as early as seen in our study compared to Group BC from minute 150 (P < 0.05). In Group [4, 5]. Keshaviah et al [4] observed a decrease of SVR and of

1173

Herrero eta!: Effects of buffer in hemodialysis

Table 2. Analytical data

A 6

Acetate

5 L 0

E E

4 3

G)

a>

0

2 1

0 0

60

120

180

KT/V Sodium mmol/liter Minute 0 Minute 180 Osmolality mOsm/kg H20 Minute 0 Minute 180 Calcium mg/dl Minute 0 Minute 180 Potassium mmol/liter Minute 0 Minute 180 Hematocrit % Minute 0 Minute 180

1.01

Acetate + Lactate Bicarbonate

0.06

1.02

0.03

1.11

0.07

138.0

1.0

136.3

0.32

137.4 136.2

0.88 0.36

137.1 136.1

0.23

335.1

3.71

330.9

5.35

330.6 299.1

4.46 2.66"

301.7 309b 301.8 3.65" 8.72

1.18

9.10

0.26

8.66

0.45

0.45

4.88

0.35

5.61

0.51

2.57 2.59

33.81 34.21

2.15 2.30

35.90 36.14

2.47 2.80

10.41

0.21 0.20"

5.00 36.95 38.42

3.28 0.19

10.33 0.20

3.34 0.18

9.92 027b

3.50 0.12

Data are means SEM. Minute 0 represents the baseline values before the hemodialysis. P < 0.01; "P < 0.001 statistical significance of values at minute 0 versus

3.5

minute 180. None of the both baseline and minute 180 values differed significantly from each other for the three hemodialysis.

B

3

I

2.5 -

0 E E a) Ce

Table 3. Gasometric baseline values

PH PaCO2 mm Hg HCO> mmol/liter Pa02 mm Hg P(A-a)02 mm Hg C(a-v)02 vol% DO> ml 02/min/m2 V02 ml 0>/minIm2 QsIQt %

2

1.5 1

0.5

0 0

60

120

180

Time, minutes Fig. 5. Blood acetate and L-lactate levels. Symbols are as in Figure 1.

Acetate

Acetate+Lactate

Bicarbonate

7.260 0.02

7.222 0.02

7.249 0.04

35.40

35.64

1.90

15.56 0.28 167.0 7.27 58.73 7.43 2.33 0.22 951.8

155.1

118.73 14.34

15.70 1.77

1.51

14.69 0.37 153.3 8.14 72.28 8.20 2.62 0.27 862.3 88.6 136.61 14.95

15.35

1.87

37.08 16.38 177.2 46.88 2.97 842.5 141.01

2.35 1.11 12.01

11.79

0.46 103.6 19.52

11.80 2.97

Data are means SEM. None of the baseline values differed significantly from each other. Abbreviations are: PaCO2, Partial pressure of arterial carbon dioxide; Pa02, Partial pressure of arterial oxygen; P(A-a)02, Alveolo-arterial oxygen pressure difference; C(a-v)O>, Arterio-venous oxygen content difference; DO>, Tissue oxygen supply; VO>, Oxygen consumption; QsIQt, Right-to-left shunting.

After the first few minutes, the evolution of the hemodynamic blood pressure in the first minute of acetate infusion in nonuremic, anesthetized dogs. Daugirdas, Nawad, and Hayashi [5] parameters showed marked differences between the groups. also observed peripheral vasodilation in the first minute of HD While in Groups AC+LA and BC there were no significant with acetate in non-uremic dogs. In both studies, after a few changes, in Group AC there was a decrease of CI, SI and MBP, minutes there was a "rebound" effect of SVR recovery, which was with a reduction of the LVSWI only in this group. The cardiac complete when the animals had been previously subjected to performance is dependent on the preload, the afterload, the heart chemical sympathectomy [5]. The mechanisms of this recovery are

rate and the contractile capacity of the myocardium [31]. In our

unknown, but the authors speculate that increased levels of case, and as regards Group AC, we can rule out changes in the vasoactive hormones or metabolic adaptation to the action of preload, as neither the CVP (which expresses the end-diastolic acetate might be some of the implicated mechanisms [5]. A filling pressure of the right ventricle) nor the PCWP (which recovery of SVR was also observed in our study, and as is the case expresses that of the left ventricle) varied during the study. in other reports [4, 5], the mechanism is unknown. In our study, Similarly, an increase of the afterload (which would have a the smaller decrease of SVR in the first minute of HD in Group

negative influence on myocardial contractility) can also be ruled

AC+LA as compared to Group AC could be due, on the one out, as the SVR were not different from the baseline ones in these hand, to the smaller amount of acetate provided in the former, stages of HD or even showed a trend to decrease. As neither the and on the other hand to a lesser vasodilator effect of lactate as preload nor the afterload varied, the decrease in SI in Group AC compared to acetate [30]. In both groups, AC and AC+LA, there was an increase in heart rate and cardiac index in the first minute

must have been due to a reduction of the myocardial contractility together with the increased heart rate. In the absence of compen-

of HD; both these increases are physiologic mechanisms of satory vasoconstriction, the increase in HR would be the only compensation of the vasodilator effect [3—6].

available mechanism to maintain cardiac output against a negative

1174

Herrero eta!: Effects of buffer in hemodialysis

A 10

*** 8 Q)

0 E E

6' 0 I

***

T-

6

***

I-- -

*** L--"

4

*** ***

5*,-

2

211:1

0 —2

05 15 30 45 60

**

4

0

dilator effect arose during the formation of AMP. The same authors, in another study, [33] postulated that the myocardial

--

---

E

6 0

**

**

6

2

120 150 180

B

8

E

90

in the flattening of the ventricular function curves, that is, in the loss of the correlation between LVSWI and PCWP during dialysis. The reduction of the SI/PCWP and LVSWI/PCWP ratios in this group further indicates a reduction of myocardial contractility [6]. In the present study, the negative correlation between plasma acetate levels and the LVSWIIPCWP ratio in Group AC suggests a negative effect of acetate on myocardial contractility during the dialysis session. The negative relationship at minute 60 did not increase at minutes 120 and 180, perhaps because the depressor effect of acetate was already near its maximal activity. In vitro studies have shown that acetate has a vasodilating [30, 32] and a myocardial depressant [33] effect, both of them at concentrations normally achieved in HD. Both the vasodilation and the myocardial depression could arise as a consequence of the formation of intermediate metabolites. This so-called metabolic theory of the acetate action was first proposed by Liang and Lowenstein [3] who, during acetate infusion in dogs, observed increased tissue levels of adenosine and adenosine monophosphate (AIVIP), and postulated that vasodilation could be due to the action of adenosine generated from AMP by the action of the enzyme 5-nudeotidase. Daugirdas and Nawad [32] have reported that the vaso-

depressant effect of acetate could be due to the action of

aT T TT

T

I

I

adenosine, as this has been shown to have a negative inotropic effect on tissue samples isolated from the atria or the papillary muscles [34]. It is possible that, in our own study, no negative inotropic effect was observed in Group AC+LA because both the

amount of acetate provided and the plasma acetate levels

0-

achieved were low. However, we do not know the possible role of lactate, to which both cardiostimulating [35] and cardiodepressant [36] actions have been attributed.

4. —6

05 15 30 45 60

90

120 150 180

In the present study, only Group AC showed a reduction in Pa02. The main mechanism of hypoxemia caused by the use of acetate in the dialysis fluid is alveolar hypoventilation [15—17]. In our case, the fall in Pa02 in Group AC occurred in the absence of

0.2

C

hypoventilation, as the animals were under constant mechanical ventilation, Studies performed both in humans [37] and in dogs [38], also under general anesthesia and mechanical ventilation, have failed to show changes in Pa02 during HD with acetate. In

0.15

contrast, other authors report reductions in Pa02 in patients

Ia.

dialyzed with acetate under mechanical ventilation [39, 40]. The causes of these discrepancies are probably to be found in differences in the populations studied, in the biocompatibility of the membranes used, in the efficiency of the dialysis (and thus in the

0.1

0.05

amount of acetate provided), and in the hemodynamic performance during HD. In our study we used a biocompatible mem-

0 —0.05

brane which has been shown to have no influence on Pa02 nor on pulmonary hemodynamics [15, 41]; furthermore, the anesthetic procedure was the same for all three groups. It could be surmised

0 5 15 30 45 60

90

120 150 180

Time, minutes Fig. 6. Arterial acid-base balance values, PaCO2 dioxide pressure. Symbols are as in Figure 1.

is arterial partial carbon

that, in Group AC, adenosine derived from the metabolism of acetate might have been the cause of the drop in Pa02, either through its bronchoconstrictive action [42] or through an alteration of pulmonary hemodynamics [43]. We did not measure pulmonary airflow resistances, although indirect data suggest that

bronchoconstriction, if at all present, must have been slight, as inotropic effect, as already reported [9, 31]. The blood pressure bronchoconstriction increases the intrapulmonary shunt [29], and decrease in Group AC after the initial minutes of the study was this was actually decreased. There were also no changes in PVR. caused by the decrease in cardiac output without variations in The increase in MPAP only during the first five minutes of HD in SVR or CVP. The myocardial depression in this group is evident groups AC and AC+LA might have been a consequence of the

Herrero et al: Effects of buffer in hemodialysis

30 20 10

20 15

1 c

I

I

_

-10

- —10

-20

—15 —20

—30 —40

0 o

**

<0

I

05 15 30 45 60 90 120 150 180

40 30

1175

** ** **

05 153045 60 90 120 150 180

60 50 40

20

10 —10

-20

0 ic

>0

I

—30 —40

—10 —20

05 15 30 45 60 90 120 150 180 120 100

05 15 30 45 60 90 120 150 180

5

80 60

40

0

20

0

.LT i

I

—5

—20 —40

05 15 30 45 60 90 120 150 180

05 15 30 45 60 90 120 150 180

Time, minutes

Time, minutes

increased cardiac output, itself a consequence of peripheral vasodilation [51. In Group AC an increase in P(A-a)02 was seen,

suggesting a ventilation/perfusion impairment, together with and increase in C(a-v)02, both of which may be caused by a decrease

in cardiac output [29]. A reduction of Pv02 appears when the

Fig. 7. Oxygenation parameters. Abbreviations are: Pa02, arterial partial oxygen pressure; P(A-a)02, alveolo-arterial oxygen pressure difference; C(a-v)02, arterio-venous oxygen content difference; Qs/Qt, right-to-left shunt; DO2, tissue oxygen supply; V02, oxygen consumption. Symbols are as in Figure 1.

Some authors have reported that changes in the acid-base balance during HD with acetate might act as adjuvant to the latter's effect as a depressant of myocardial contractility [11]. Both a worsening of metabolic acidosis [1] and hypocapnia [15—17] may occur during HD with acetate, and both have a negative inotropic effect [46]. In view of constant mechanical ventilation, the PaCO2 fall in Group AC must be a consequence either of CO2 loss into

tissue oxygen extraction increases, and this happens in such widely different situations as sudden anemia, increased metabolism with increase of oxygen uptake, and reduction of cardiac output [29]. dialysate [15, 16] or of a lower CO2 generation during acetate In our study, anemia did not arise in any of the groups, nor were metabolism [17]; in this group, we cannot rule out that the there significant changes in oxygen consumption. Based on these reduction of PaCO2 might have negatively influenced myocardial

data, we consider that the Pa02 decrease in Group AC was due primordially to a fall in cardiac output, as it is well-known that a reduction in this parameter typically induces a slight drop in Pa02, because when the cardiac output decreases, the shunted blood entering the arterial circulation will have a lower Pv02 than while it was normal [29]. Changes in sodium and osmolality [44], as well as in potassium

contractility, although the PaCO2 reduction, though significant, was slight. The PaCO2 did not change in Group AC+LA, even though the CO2 losses into the dialysate were probably similar to those seen in Group AC. This finding agrees with a previous observation in patients with chronic renal failure when a dialysis fluid with the same composition was used [24]. Several factors

might have been at the origin of the divergent evolution of PaCO2 in Groups AC and AC+LA. In the first place, it is well-known that In the present study, however, it is improbable that variations in CO2 production during metabolism of lactate is twice that seen these parameters may have had an influence on the reduction of during metabolism of acetate [35, 47, 48]; a further possibly myocardial contractility in Group AC, as there were no significant contributing factor may have been bicarbonate hydrolysis, as the and calcium [45], may modify myocardial contractility during HD.

changes in the sodium levels, and the decreases in plasma rise of the blood bicarbonate levels occurred earlier and, even osmolality and plasma potassium, and the rise in the calcium levels were similar in all three groups.

though it did not achieve statistical significance, tended to be slightly higher in Group AC+LA.

1176

Herrero eta!: Effects of buffer in hemoditdysis

With the use of AC+LA concentrate the plasma levels both of acetate and of L-lactate were low. Acetate is metabolized mainly in striated muscle and to a lesser proportion in the liver [47], while

lactate is metabolized to a greater extent in the liver, and to a lesser one in muscle [48]; furthermore, the metabolic pathways are different for both anions [47, 48]. This renders it possible to use simultaneously acetate and lactate in hemodialysis fluids as, within the organism, they will be transformed complementarily through metabolic pathways and in tissues for neutralization of hydrogenions. In theory, the acetate-lactate mixture would offer advantages over acetate as single buffer when, with the latter, the

7. Aizw Y, OHM0RI T, LrIA1 K, NA1t Y, MAT5UDKA M, HIILA5AwA Y:

Depressant action of acetate upon the human cardiovascular system. Clin Nephrol 8:477—480, 1977 8. KIRKENDOL PL, PE.&R5ON JE, BOWER JD, HOLBERT RD: Myocardial

depressant effects of sodium acetate. Cardiovasc Res 12:127—136, 1978 9. VINCENT JL, VANHERWEGHEM JL, DEGAUTE JP, BERRE J, DI.JFAYE P,

KAHN RJ: Acetate-induced myoeardial depression hemodialysis for acute renal failure. Kidney Int 22:653—657, 1982 10. RUDER MA, ALPERT MA, Viw STONE J, SELMON MR, KELLY DL, 11AYNIE JD, PERKINS SK: Comparative effects of acetate and bicar-

bonate hemodialysis on left ventricular function. Kidney Int 27:768— 773, 1985 11. LEUNIsSEN KML, HOORNTJE SJ, FmRs HA, DEKKER5 WT, MULDER

maximal rate of acetate metabolism is widely exceeded; this

AW: Acetate versus bicarbonate hemodialysis in critically ill patients.

occurs mainly in high-efficiency HD and/or with patients having a

12. MAN5ELL MA, MORGAN SH, MOORE R, KONG CH, LAKER MF, WtNG

Nephron 42:146—151, 1986

low muscular mass [47]. When this happens, a progressive inA.J: Cardiovascular and acid-base effects of acetate and bicarbonate crease of the plasma acetate levels is seen during the HD session, haemodialysis. Nephrol Dial Transplant 1:229—232, 1987 while at the same time the metabolism of the anion insufficiently 13. NIXON JV, Mrrcanu. JH, McPI-t.&uL JJ, HENRICH WL: Effect of hemodialysis on left ventricular function. Dissociation of changes in compensates the bicarbonate losses into the dialysate [47]. We filling volume and in contractile state. J C/in Invest 71:377—384, 1983 think that, in our study, the maximal rate of acetate metabolism 14. WIZEMANN V, SOETANTO R, THORMANN J, LUEBBECKE F, KRAMER W: was only exceeded in Group AC. This would explain the differEffects of acetate on left ventricular function in hemodialysis patients. Nephron 64:101—105, 1993 ences in the plasma acetate levels between Group AC and Group AC+LA, as well as the evolution of the blood bicarbonate levels, 15. D BACKER WA, VERPOOTEN GA, BORGONJON DJ, VERNEIRE PA, LIN5 RR, DE BR0E ME: Hypoxemia during hemodialysis: Effects of although it is well-known that lactate conversion into bicarbonate different membranes and dialysate compositions. Kidney Int 23:738— is slower than that of acetate [23, 35]. 743, 1983 In summary, even though we acknowledge that this experimental model cannot be fully extrapolated to chronic patients on HD, we have ascertained that acetate as the single buffer may induce hemodynamic instability through peripheral vasodilation and de-

pression of myocardial contractility. Even in the absence of

16. Do MJ, WHwP BJ, DAvIDsoN WD, WEIrZMAN RE, WASSERMAN K: Hypopnea associated With acetate hemodialysis: Carbon dioxide flow-dependent ventilation. N EngI J Med 305:72—75, 1981 17. OH MS, URIBARRI J, DEL Mowit ML, HENEGHAN WF, KnE CS, FRIEDMAN EA: A mechanism of hypoxemia during hemodialysis. Consumption of CO2 in metabolism of acetate. Am J Nephrol 5:366— 371, 1985

hypoventilation, acetate may cause a fall in the arterial oxygen pressure, through a reduction in cardiac output. The equipropor- 18. EI5ER AR, JAYAMANNE D, KORSENG C, CHE H, SLWKIN RF, NErs MS: Contrasting alterations in pulmonary gas exchange during acetate and tional mixture of acetate and lactate also induces vasodilation, but bicarbonate hemodialysis. Am J Nephrol 2:123—127, 1982 demonstrates advantages over acetate as single buffer in that it 19. GOTCH FA: Dialysis of the future. Kidney Int 33(Suppl 24):S-l00—Sdoes not cause myocardial depression and thus does not modilS' 104, 1988 20. GEERLING5 W, Tut'vEsoN 0, EHRICH JHH, JONES EHP, LANDAI5 P, the arterial oxygen pressure through this mechanism. Acknowledgments This study was partly supported by Grant F.I.S. 90/106 from spanish Fondo de Investigaciones Sanitarias de Ia Scguridad Social. The authors thank Dr. MC. Ruiz and Dr. D. Hernandez from Clinical Biochemistry Department of Hospital Universitario San Carlos, Madrid, for technical assistance. This manuscript was presented in part at the XXIX Congress of the European Dialysis and Transplant Association-European Renal Association, June 1992. Reprint requests to Dr. Jose A. Herrero, Servicio de Nefrologia, Hospital Universitario San Carlos, Martin Lagos s/n, 28040 Madrid, Spain.

LoIRAT C, LOIRAT C, MALLICK NP, MARGREITER R, RAINE AEG, SALHELA K, SELW00D NH, VALDERRABANO F: Report on manage-

ment of renal failure in Europe, XXIII, 1992. Nephrol Dial Transplant 9(Suppl 1):6—25, 1994 21. YAMAUCHI M, KUMAGAI Y, KAMATA T, EzUMI A, SHon R: Clinical

effects of sodium lactate dialysate on acetate intolerance. Nippon Jinzo Gakkai Shi 26:63—69, 1984 22. NAWAB ZM, ARMSTRONG MK, WEISSBERGER LE, ING TS, HAYA5HI

JA, DAUGIRDAS JT: Use of lactate as a base in hemodialysis. Am J Nephrol 7:434—439, 1987

23. DMS 5, Yti AW, GUpTA DK, KAR PM, ING TS, DAUGIRDAS iT: L-lactate high-efficiency hemodialysis: Hemodynamics, blood gas changes, potassium/phosphorus, and symptoms. Kidney Int 38:896— 903, 1990 24. TORRENTE J, CORONEL F, HERRERO JA, MACIA M, BARRIENTOS A:

References 1. NOvELLO A, Kasci-i RC, EA5TERLINO RE: Acetate intolerance during hemodialysis. din Nephrol 5:29—33, 1976 2. HAMPL H, Kinpp H, WOLFGRUBER M, PU5TELNIK A, SCHILLER R, HANEFE5LD F, KE55EL M: Advantages of bicarbonate hemodialysis. mt JArtif Organs 6:410—414, 1982

3. LIANG CS, LownsNsTEIN JM: Metabolic control of the circulation. Effects of acetate and pyruvate. J dim Invest 62:1029—1038, 1978 4. KE5HAvIAH P, CAIu.soN L, C0N5mNTINI E, SRkPIR0 F: Dialysis induced hypoxemia and hypotension are not casually related. ASAIO Trans 30:159—162, 1984

5. Dis.uornas.s JT, NAwAD ZM, HAYA5HI JA: Hemodialysis hemody-

namics in an animal model: Effect of using an acetate-buffered dialysate. J Lab Clips Med 107:517—524, 1986

6. ScHoI* D, ScHMITI' R, JAHN H: Myocardial impairment in chronic hemodialysis patients. Contrib Nephrol 52:69—85, 1986

Partial substitution of sodium lactate for sodium acetate in the bath fluid for hemodialysis. Artif Organs 14:2—6, 1990 25. MOFFATr JD, BALLANTYNE WC: Monitoring techniques and normal values, in Synopsis of Critical Care, edited by SIBBALD WJ, Baltimore, Williams Wilkins, 1984, pp. 309—3 18 26. BRAUNWALD E: Assessment of cardiac function, in Heart Disease (3rd ed), edited by BRAUNwALD E, Philadelphia, WB Saunders Company, 1988, pp. 449—470 earboxili27. RUIz MC: Dosificación eromatográfica en sangre de eos de eadena corta. Biometrica 10:17—22, 1985 28. GUTMAN I, WAHLFELD AW: L-(+)-laetate determination with lactate

dehydrogenase and NAD, in Methods enzymatic analysis (2nd ed), edited by BERGMEYER HU, New York, Academic Press mc, 1974, pp. 1446—1451

29. MALLEY WJ: Clinical Blood Gases: Application and Noninvasive Alternatives, Philadelphia, WB Saunders Company, 1990 30. NAWAB ZM, DAUGIRDA5 JT, ING TS, LEEHEY DJ, REID RW, KL0K

Herrero eta!: Effects of buffer in hemodialysis

MA: Calcium-complexing versus vasorelaxant effects of acetate, lactate and other bases. ASAIO Trans 30:184—188, 1984 31. BRAUNWALD E, SONNENBLICK EH, Ross J: Mechanisms of cardiac

contraction and relaxation, in Heart Disease (3rd ed), edited by BRAUNWALD E, Philadelphia, WB Saunders Company, 1988, pp. 383—425

32. DAUGIRDAS JT, NAWAB ZM: Acetate relaxation of isolated vascular smooth muscle. Kidney mt 32:39—46, 1987 33. DAUGIRDAS JT, WANG X, NUTrING C, SWANSON V, AGRAWAL A: Depressant effect of acetate in isolated cardiac tissue. Cardiovasc Res 22:566—570, 1988

34. BOHN M: Cardiac effects of adenosine. Mechanism of action, pathophysiologic and clinical significance. Kim Wochenschr 65:487—499, 1987 35. Br.s.siou S, CHIARAMONTE S, FABRIS A, FERIANI M, PIsANI E, RONCO

C, LA GRECA G: Acid base balance in peritoneal dialysis using acetate and lactate, in UremicAcidosis and Extracorporeal Treatment, edited by PETRELLA E, Milan, Wichtig Editore, 1983, pp. 97—122

36. SAMAN S, Ore LH: Mechanism of reduction of action potential duration of ventricular myocardium by exogenous lactate. J Mo! Cell Cardiol 10:659—662, 1984 37. BOUFFARD Y, VIALE JP, ANNAT G, GUILLAUME C, PERcIvAL C, BERTRAND 0, MOTIN J: Pulmonary gas exchange during hemodialysis. Kidney mt 30:920—923, 1986 38. RALPH DD, OTr SM, SHERRARD Di, HIASTAIA MP: Inert gas analysis

of ventilation-perfusion matching during hemodialysis. I C/in Invest 73:1385—1391, 1984 39. CARLON GC, CAMPFIELD PB, GOLDINER PL, TURNBULL AD: Hypox-

emia during hemodialysis. Crit Care Med 7:497—499, 1979 40. JONES RH, BROADFIELD JB, PARSONS V: Arterial hypoxemia during

1177

hemodialysis for acute renal failure in mechanically ventilated patients: Observations and mechanisms. Cliii Nephroi 14:18—22, 1980 41. DEL CASTILLO D, GUERRERO R, BARRIO A, ESCASSI A, MARTIN-

MAL0 A, ALJAMA P: Hemodinámica pulmonar e intercambio alveo-

locapilar durante Ia hemodiálisis: Estudio en un modelo animal. NefrologIa 9:293—303, 1989 42. CHURCH MK, HOLGATE ST: Adenosine-induced bronchoconstriction

and its inhibition by nedocromil sodium. J Alleigy C/in Immunol 92:190—194, 1993 43. NEELY CF, KADOWITZ PJ, LLPPTON H, NEIMAN M, HYMAN AL:

Adenosine does not mediate the pulmonary vasodilator response of

adenosine 5'-triphosphate in the feline pulmonary vascular bed. I Pharmaco/ Exp Ther 250:170—176, 1989 44. KOZENY GA, MURDOCK DK, EULER DE, HANo JE, SCANLON PJ,

BANSAs. VK, VERTUNO LL: In vivo effects of acute changes in osmolality and sodium concentration on myocardial contractility. Am Heart J 109:290—296, 1984 45. HENRICH WI, HUNT JM, NIXON JV: Increased ionized calcium and left

ventricular contractility during hemodialysis. NEnglJMed 310:19—23, 1984 46. MOLONY DA, SCHIESS MC, EvAN0FF GV: Respiratory acid-base disorders, in Fluids and Electrolytes (2nd ed), edited by KOKKO JP, TANNEN RL, Philadelphia, WB Saunders Company, 1990, pp. 391—484 47. VINAY P, C.Doso M, TEJEDOR A, PRUD'HOME M, LEVELILLE M, VINET B, COURTEAU M, GOUGOUX A, RENGEL M, LAPIERRE L, PIETFE Y: Acetate metabolism during hemodialysis: Metabolic considerations. Am I Nephroi 7:337—354, 1987 48. COHEN RD, SIMPSON R: Lactate metabolism. Anesthesiology 43:661— 673, 1975