Priming Solutions

for Cardiopulmonary

Bypass: Comparison

of Three Colloids

D. Himpe, MD, P. Van Cauwelaer-t, MD, H. Neels, PhD, D. Stinkens, RN, F. Van Den Fonteyne, RN, W. Theunissen, MD, P. Muylaert, MD, C. Hermans, MD, G. Goossens, MD, J. Moeskops, MS, J. Van Hoof, MD, J. Alleman, MD, and H. Adriaensen, The present study was designed to compare the differences in the clinical effects of three colloidal solutions, albumin, urea-linked gelatin, and succinyl-linked gelatin, when used as priming fluids for cardiopulmonary bypass (CPB) under *stat conditions. A consecutive series of 105 patients scheduled for cardiac surgery were randomized into three identically managed groups, except for the CPB prime. Variables relating to acid-base status, oncotic activity, metabolism, coagulation, and postoperative evaluation were measured. Marked differences in acid-base status, colloid osmotic pressure, additional prime requirements, blood lactate, urine output, and the need for buffer solutions occurred among groups, with the succinyl-linked gelatin group having better results

S

INCE BLOOD-FREE primes for cardiopulmonary bypass (CPB) were introduced during the early days of adult cardiac surgery, various kinds of intravenous (IV) solutions have been used for hemodilution.’ However, no consensus has been reached on the composition of the ideal prime for CPB. Priming solutions may influence the physicochemical and homeostatic equilibria in the blood and modify the metabolic response to cardiac surgery using CPB.’ Furthermore, acid-base management during hypothermic CPB, whether pH-stat or a-stat,’ could be influenced by the priming fluid selected.4 During hemodilution, excessive fluid shifts to the interstitial space may occur and are frequently associated with respiratory and cardiac dysfunction, impairment of wound healing, and decreased resistance to infection in the postoperative period.’ According to Starling’s hypothesis, this can be avoided by the addition of a colloid. However, the choice between the use of colloids or crystalloids for hemodilution during CPB has long been a matter of controversy.6 The proponents of crystalloids argue that these solutions are inexpensive, nontoxic, and free from the risk of anaphylactoid reactions.’ Assuming that relative hypovolemia and hypotension during CPB lower mean capillary pressure, they also claim that an important decrease in colloid osmotic pressure (COP) can be permitted without the movement of extra fluid into the interstitium.8 Other investigators found important fluid shifts, but continue to argue that no positive effects of colloid in the prime can be shown except in selected cases9 However, detrimental consequences of interstitial fluid gain are well recognized today,“‘~” and there is growing evidence that it is advantageous to replace crystalloids with colloids.” On the basis of these theoretical considerations and practical experience, colloids chosen for CPB must have sufficient and prolonged oncotic activity, good biocompatibility with no interference with the coagulation or any other system, limited survival in the body, minimal risk for adverse reactions and deleterious effects, and a reasonable cost. Gelatins are processed by the degradation of collagen. Cleavage of the molecular polypeptide chains and crosslinking them with different chemical substances, such as glyoxal, succinic acid, or isocyanide, result in three different Journalof Cardiothoracic and Vascular Anesthesia,

MD, PhD

than the other groups. Changes in hemodynamics, oxygen consumption, and blood-glucose levels during CPB did not vary among groups. There were also no important intergroup differences in hematologic and clotting variables or postoperative parameters such as blood loss or use of blood products. Electrolyte changes were similar except for a significant increase in ionized calcium that occurred in the urea-linked gelatin group after bypass. The results indicate that succinyllinked gelatin is an adequate and safe alternative to human albumin for use as a colloid during CPB under q-stat conditions. Copyright o 1991 by W.B. Saunders Company

classes of polypeptides: the oldest preparation oxypolypeptide, and the newer succinylated and urea-cross-linked polypeptides, respectively. Therefore, artificial colloids are polydispersed, which may confer some important advantages when a colloid, such as the gelatins, has a relatively small molecular weight (M, = 30,000 to 35,000). For example, the smaller molecules may initially reduce blood viscosity and promote diuresis, whereas larger molecules may prolong plasma expansion. However, synthetic colloids with more elevated molecular weights increase plasma viscosity” and may impair kidney function by “plugging” of the tubules.14 The lower molecular weight of the gelatins further results in a greater water-retaining capacity because more osmotically active particles are present for a given concentration, although the duration is shorter because of the more rapid elimination without storage in the reticuloendothelial system.” Gelatins are free of adverse effects on hemostasis, cross-matching, and renal function; thus, no dose limit is recommended.‘6~‘7 The newer starch analog, pentastarch, which was not available for clinical use at the time of this study, may eventually be proven to have less effect on blood coagulation than other starch derivatives. Preliminary results demonstrate some properties similar to those described previously for the gelatins.18B’9 Few data are available on gelatins used as priming solutions based on oncotic performance differences, and there are no data on their influence on acid-base management during CPB.20,21 Therefore, this prospective, randomized study selecting two new gelatins (urea-linked and succinyl-linked) for

From the Departments of Anesthesiology and Cardiac Surgery, the Perfusion Team, and the Laboratories of Clinical Chemistry and Hematology, Middelham General Hospital, Antwerp Belgium; the Institut Merieux Benelux, Brussels, Belgium; and the Department of Anesfhesiology, University ofAntwerp, Antwerp, Belgium. Address reprint requests to D. Himpe, MD, Department of Anesthesiology, Middelheim General Hospital, Lindendreef 1, B-2020 Antwerp, Belgium. Copyright 0 1991 by W.B. Saunders Company 1053-0770/91/0505-0008$03.00/0

Vol5, No 5 (October), 1991:

pp457-466

457

458

comparison with concentration in as the priming revascularization

the “standard colloid” albumin at a similar lactated Ringer’s or Hartmann‘s solution fluid for CPB during coronary artery under a-stat conditions was conducted.

MATERIALS

AND METHODS

After protocol approval by the Institutional Review Board, informed consent was obtained from all participants. During a 3-month period, all men aged 40 to 70 years scheduled for elective coronary artery surgery (3 to 4 coronary artery bypass grafts) entered the study. This provided a consecutive series of 105 subjects. Patients with significant coagulopathy, anemia, chronic renal failure, respiratory insufficiency, diabetes mellitus, or liver dysfunction were excluded. The treatment for all patients was identical, except for the composition of the priming fluid. Immediately prior to surgery, patients were randomly allocated to one of the three groups using a random number table. A volume of 2,200 mL was necessary to prime the oxygenator (CMSO; Bentley, Irvine, CA) and extracorporeal circuit. The composition of the primes in the three groups was as follows: group I (Hartmann’s solution + 20% albumin) contained 130 mEq Na’. 4 mEq K’, 109 mEq CT, 3 mEq Ca+‘, 0 mEq Mg”, 28 mEq lactate. and 2.7% albumin; group II (polygeline or urea-linked gelatin, Haemaccel) had 145 mEq Na”, 5.1 mEq K’, 145 mEq Cl-. 12.5 mEq Ca+‘, 0 mEq Mg”, 0 mEq lactate, and 3.5% gelatin: and group III (modified fluid gelatin or succinyl-linked gelatin, Geloplasma) had 150 mEq Na+, 5 mEq K’, 100 mEq Cll, 0 mEq Ca++. 3 mEq Mg++. 30 mEq lactate, and 3% gelatin. The standard anesthetic approach consisted of administering large doses of fentanyl, 50 to 100 kg/kg, pancuronium, 0.1 to 0.2 mgikg, and flunitrazepam, 0.5 to 2 mg, with a dose of lidoflazine, 1 mgikg, which is routinely administered in this institution for myocardial” and cerebral” protection. All patients received 2 g of methylprednisolone, 6 g of piracetam, and 20 mg of furosemide before CPB. Heparin and protamine administration was managed by repeated measurements of the activated clotting time and the use of a computerized, two-point dose-response method.‘” Distal anastomoses were performed on the fibrillating heart using the intermittent cross-clamping technique without administration of cardioplegic solutionsZi Acid-base status was managed following the o-stat concept while the patient’s blood was cooled to 30°C. Arterial pH and CO, tension were maintained in the “normal” 37°C ranges, the samples being measured at the analyzer temperature (37°C) without correction for the actual temperature. However, for safety, arterial PO? was corrected to the actual temperature and maintained between 100 and 200 mm Hg. Sodium bicarbonate was administered if the base deficit exceeded 5 mEq/L. At the end of CPB, just prior to separation, 1,375 mg of calcium glubionate was administered. Arbitrarily, mean perfusion pressures were maintained between 50 and 75 mm Hg. Flow rates of 2.4 ? 0.4 Liminim’were maintained even during hypothermia. Hypotension was treated with a-agonists (eg, phenylephrine), and elevated pressure\ were lowered by isoflurane delivery via the oxygenator. Preoperatively, the body surface area was determined, and the following hematologic and clotting variables were measured: hemoglobin, hematocrit, activated partial thromboplastin time (aPTT), prothrombin time (PTT quick, % of normal), fibrinogen concentration, and platelet count. Kidney function was evaluated by assessing creatinine blood levels. After induction of anesthesia and prior to surgery, arterial blood gas analysis was performed, and the serum sodium, ionized calcium, potassium. chloride, and hemoglobin were determined. After heparinization before CPB, baseline levels of COP and blood

lactate were measured together with the hemoglobin IC\C! .iiic platelet count. The COP was measured with a membrane colloiil CimhIl. Berlin osmometer (BMT-921: BMT MESSTECHNIK Germany) fitted with a semipermeable membrane with a cut-crtt’for molecules above IO.000 d, Arterial and venous blood gas analyses to calculate oxygen consumption. hemoglobin, platelet count, COP, and lactate level\ were repeated every 15 minutes during hypothermic CPB and once after rewarming to normothermia. Within 20 minutes after prot;tmine administration. measurements were made of hemoglobin. platelet count, free plasma hemoglobin. COP, blood lactate. sodium, ionized calcium. potassium. chloride. aPTT, PTT. and fibrinogen concentration. Except for free plasma hemoglobin. these measurements were repeated on arrival of the patient in the intensive care unit (IO). Six hours later: and on the morning ol the first postoperative day, the total protein, platelet count. aPTT. PTT, and fibrinogen level were determined. The total protein concentration in blood, platelet count, and PTT were repeated 4X hours after arrival in the ICIJ. Other intraoperative data recorded for each patient included duration of CPB. urine production, requirements for additional prime to maintain pump volume, and frequency of sodium bicarbonate administration to correct met+ bolic acidosis. The mean arterial blood pressure and pump fou were registered every 15 minutes for subsequent calculation of the systemic vascular resistance (SVR). During the patient’s stay in the ICU, therapy was directed by the on-duty staff without regard to the investigation. The administration of blood and blood products was guided by clinical assessment of the patient’s general appearance, blood loss, blood pressure. tilling pressures. urine production. and blood analyses. Interstitial fluid accumulation waj estimated by the ICU physicians using chest x-rays.” Hypovolemia was always corrected by stable solution of plasma proteins (SSPP). Differences in postoperative administration of blood products were evaluated by comparing the respective frequencies of a I-U transfusion in each of the groups for packed cells. fresh frozen plasma (FFP). SSPP. and concentrated human albumin. Daily measurements of creatinine blood levels acre performed to evaluate kidney function. On patient discharge from the 10, the following clinical observations were recorded: total amount of chest tube drainage. hematocrit. signs of allergic reactions, need for prolonged intubation (> 24 hours), duration of ICIJ stay, and necessity for reoperation for bleeding. Table 1 lists the time codes used throughout the study. The overall significance level was fixed at 0.05 for the statistical analysis to compare the effects of one factor (priming fluid) on several variables. Bonferroni’s correction was chosen to compensate for multiple comparisons keeping the type I error at the preset level. Nonparametric data were analyzed with the x’ test with

Table 1. Time Points in the Study Time

Point

Preoperative value

t-1

Sample obtained just prior to

to tl

through tll

CPB

Samples obtained during CPB under hypothermia (30°C) every 15 minutes

t12

Last sample during

t13

Sample 20 minutes after protamine administration

t14

Sample on arrival at the ICU

CPB

after rewarming to 37°C

t15

Sample 6 hours after ICU admission

t16

Sample 24 hours after ICU admission

t17

Sample 48 hours after ICU admission

t18

Sample on discharge from the KU

COMPARISON OF THREE COLLOIDS

459

Table 2. Patient Profiles

contingency tables and, when required, with Fischer’s exact test. Parametric data are expressed as mean f SEM. Baseline group comparisons and individual measurements were analyzed by a or by a one-way analysis of variance (ANOVA) when appropriate

Group I

II

Ill

Age (vr)

60 + 1.3

57 f 1.6

60 2 1.2

Weight (kg)

73 + 1.1

73 + 1.7

75 + 3.0

1.84 + 0.01

1.85 + 0.02

1.82 2 0.02

3 (2-4)

3 (2-4)

3 (2-4)

BSA (m*)

Kruskal-Wallis test. Homogeneity of variances was verified using the F test. A two-way ANOVA for repeated measurements was used to detect significant differences among groups. When significant differences were present, further analysis included multiple t tests when appropriate, or Mann-Whitney U tests.

Classification Canadian Cardiovascular Society Distal anastomoses per patient Duration of CPB (min)

3.7 f 0.14

3.6 ? 0.18

3.5 + 0.11

117 + 3.9

119 f 2.8

110 ? 2.8

RESULTS

There were no significant differences among study groups in terms of clinical data (Table 2). As seen in the top curve of Fig 1, the most stable acid-base pattern was observed in group III (modified fluid gelatin, succinyl-linked gelatin). Significantly lower (P < 0.05) pH values were observed at most time points during and after bypass in group II

NOTE. All values are mean 2 SEM except for the CCS classification, which is expressed

as median (range). For all comparisons

among

groups, no statistically significant differences could be demonstrated.

A 7.5 7.48

__PH

7.46 7.44 7.42 7.4 7.30 7.36 7.34 7.32

>-------HYPOTHERMIC t0

Fig 1. Changes in acid-base equilibrium. (A) The pH values observed; (B) base excess/deficit. Values are mean + SEM. -, Group I (Hartmann-albumin); 0, group II (urea-linked gelatin); X, group Ill (succinyl-linked gelatin). ?? P < 0.05 between groups. See Table 1 for time codes.

t1

t2

t3

t4

CPB-------< t5

t6

t7

t12

t13

t14

-5

>-------HYPOTHERMIC

a t0

t1

t2

t3

t4

CPB-------< t5

t6

t7

t12

t13

t14

460

1700

HIMPE ET Ai

T

'QJ' dyne t

??

s - cm.-5

1480-1388-12aEi-IlBB-IEWJ-9n0-Fig 2. Changes in SW during CPB. Values are mean f. SEM. -, Group I (Hartmann-albumin); 0, group II (urea-linked gelatin); X, group Ill (succinyllinked gelatin). *P < 0.05 between groups. See Table 1 for time codes.

Baa-700 tl

t2

t3

t4

t5

t6

t7

(polygeline, urea-linked gelatin), as well as at several time points in group 1 (Hartmann’s solution and albumin). The evaluation of base excess/deficit levels showed a similar profile: significantly higher deficits (P < 0.05) occurred in group II as compared with those in group III (Fig 1, bottom curve). The average gas-to-blood ratios necessary to ventilate the oxygenator during CPB to maintain the preset a-stat level were 0.53 2 0.011, 0.57 ? 0.011, and 0.50 2 0.009 for groups 1, II, and III, respectively, with a significant difference noted for group II compared with groups I and III (P < 0.05, group II v I and III). The need to administer bicarbonate boluses to correct for metabolic acidosis during CPB was significantly higher (P < 0.05) in group II (13 doses) than in group III (1 dose) and group I (7 doses). Hemodynamic profiles (mean perfusion pressure, pump flow, SVR) were similar in all three groups except at the beginning of CPB, when group II had significantly higher SVR and pressure than groups I and III (Fig 2). Nonetheless, no major decreases in blood pressure ( < 30 mm Hg) occurred in any patient during initiation of

tl2

CPB or in the course of surgery. No intergroup differences in total oxygen consumption occurred during CPB (Fig 3). As expected, initial lactate levels during hypothermic CPB were significantly elevated in group III and also, but to a lesser extent, in group I due to the high lactate concentration in the two solutions (Fig 4). In group III, lactacidemia declined sharply after the initial lactate load. Lactate levels also increased with the lactate-free polygeline solution (urea-linked gelatin, group II). After 1 hour of CPB, lactate levels of the three groups were similar but remained significantly elevated over the baseline values. A significant lactic acidosis did not occur in any patient. Significant hemodilution occurred in all three groups at the start of CPB when the priming fluid was mixed with the patient’s blood. Slight initial differences were noted among the three groups, with hemodilution being the least pronounced in group I and the most pronounced in group III (hemoglobin of 9.06 ? 0.14 g/dL, 8.65 +- 0.15 g/dL, and 8.39 ? 0.15 g/dL at the start of CPB v 8.78 -C 0.15 g/dL, 8.87 of-0.17 g/dL, and 8.71 2 0.15 g/dL by the end of CPB

I

fter

T

120 Fig 3. Oxygen consumption during CPB. G&es are mian + SEM. -, G~OUD I (Hartmann-albumin); 0, group II (urea-linked gelatin); X, group Ill (succinyl-linked gelatin).

t 100

t

>--------HYPOTHERMIC

CPB--I------
-------HYPOTHERMIC 5

t0

t1

t2

for groups I, II, and III, respectively). This initial difference briefly reached statistical significance between groups I and III during the first 15 minutes of CPB (P < 0.05, group III v group I). As seen in Fig 5, significant differences in oncotic pressure were observed among groups. In contrast to an initial and transient elevation in COP, which occurred in groups II and III at the start of CPB, the oncotic pressure in group I dropped immediately. The highest COP was recorded in group III, generally remaining similar to prebypass levels. Initially, in group II, the COP levels were only slightly lower than the group III values, but declined significantly by the end of CPB. In group I, the COP slowly recovered from its initial decrease, never reaching the levels of both gelatins. During CPB, 714 ? 98 mL, 472 ? 83 mL, and 168 ? 44 mL of additional prime were required in groups I, II, and III, respectively, in order to maintain an adequate circulating volume. This difference reached statistical significance between group III and the other two groups (P < 0.05 group III v groups I and II). 25

t3

CPB-------
--------HYPOTHERMIC

CPB-------
CPB

Priming solutions for cardiopulmonary bypass: comparison of three colloids.

The present study was designed to compare the differences in the clinical effects of three colloidal solutions, albumin, urea-linked gelatin, and succ...
1MB Sizes 0 Downloads 0 Views