Tolerable infusion rate of citrate based on clinical signs and the electrocardiogram in conscious dogs

ARTICLE IN PRESS Clinical Nutrition (2006) 25, 984–993

http://intl.elsevierhealth.com/journals/clnu

ORIGINAL ARTICLE

Tolerable infusion rate of citrate based on clinical signs and the electrocardiogram in conscious dogs Tatsuru Fukuda, Shigeki Toyoshima, Yoshifumi Nakashima, Osamu Koshitani, Yoshiro Kawaguchi, Akira Momii Division of Pharmacology, Drug Safety and Metabolism, Otsuka Pharmaceutical Factory, Inc., Naruto, Tokushima 772-8601, Japan Received 5 September 2005; accepted 16 January 2006

KEYWORDS Citrate infusion; Clinical signs; Electrocardiogram; Ionized calcium; Hypocalcemia; Dog

Summary Background & aims: The possible clinical significance of the toxic effects of citrate has not yet been fully clarified. This study was therefore conducted to confirm the toxicity and determine the tolerable infusion rate of citrate administered by rapid intravenous infusion to conscious dogs. Methods: Citrate solutions were infused via the cephalic vein of 4 conscious dogs at 0.33, 0.67, or 1.33 mmol/kg/h up to 1.33 mmol/kg. Clinical signs and the electrocardiogram were observed during and after infusion. Serum citrate and ionized calcium levels were also measured. Results: Although the mean citrate level increased in accordance with the infusion rate, the calcium level decreased. No significant changes in clinical signs or the electrocardiogram were observed during infusion at 0.33 mmol/kg/h despite an increase in the serum citrate level to 1.2270.11 mmol/l (pre-infusion value: 0.3870.01 mmol/l) and a decrease in the serum calcium level to 1.2870.03 mmol/l (pre-infusion value: 1.5070.05 mmol/l). Vomiting and QTc prolongation were observed at 0.67 mmol/kg/h or higher. Salivation and tachycardia were observed at 1.33 mmol/kg/h. Conclusions: Based on clinical signs and the electrocardiogram, the tolerable infusion rate of citrate in conscious dogs is concluded to be 0.33 mmol/kg/h. & 2006 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.

Introduction Corresponding author. Tel.: +81 88 685 1151;

fax: +81 88 684 0553. E-mail address: [email protected] (T. Fukuda).

Citrate is an anticoagulant used in blood collection. It is administered during transfusion, but does not

0261-5614/$ - see front matter & 2006 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. doi:10.1016/j.clnu.2006.01.011

ARTICLE IN PRESS Tolerable infusion rate of citrate usually accumulate in the circulating blood because it is immediately metabolized in the liver. However, during rapid massive transfusion, the amount of citrate administered may exceed the metabolic capacity of the liver, and various severe toxic effects may occur. It is known that the toxic effects of citrate, including tetany, carpopedal spasm, a reduction in blood pressure, and prolongation of the QT interval that can be observed on the electrocardiogram (ECG), result from a decrease in the blood ionized calcium (Ca2+) level due to its chelating action.1–7 It is also known that the decrease in the blood Ca2+ level is related to the transfusion rate of citrated blood.6–8 Citrate has also been used as a stabilizer in commercial intravenous solutions (e.g., acetate Ringer’s solution with 1% carbohydrate and 10.5% carbohydrate solution with electrolytes). In order to employ citrate effectively in the formulation of parenteral solutions, it is necessary to determine the tolerable infusion rate that does not induce toxic effects, but this has not yet been fully clarified. The present study was therefore conducted to determine the tolerable infusion rate of citrate by observing the changes in clinical signs and the ECG and to estimate the tolerable serum citrate and Ca2+ levels when citrate was administered by rapid intravenous infusion to conscious dogs. It has also been reported that citrate chelates blood ionized magnesium (Mg2+) in patients receiving massive transfusion.9,10 Neuromuscular symptoms that mimic ionized hypocalcemia, as well as ventricular fibrillation and torsade de pointes, which occur in the setting of QT interval prolongation, are observed in patients with ionized hypomagnesemia.11,12 Therefore, the serum Mg2+ level was also measured during the course of this study.

Materials and methods Test solutions Anhydrous citric acid and trisodium citrate dihydrate (Wako Pure Chemical Industries, Ltd., Osaka, Japan) were dissolved in 200 mL of normal saline solution (Otsuka Pharmaceutical Factory, Inc., Tokushima, Japan) at a molar ratio of 1:5.25 based on CPD (citrate–phosphate–dextrose) solution (pH 5.4–5.8),13 and test solutions containing 100, 200, and 400 meq/l (33, 67, and 133 mmol/l) of citrate were prepared.

985

Animals and housing conditions Sixteen healthy male beagle dogs (Covance Research Products, Inc., Cumberland, VA, USA) weighing 9.0–12.0 kg were used after a 2-week period of acclimation to being restrained in a hanging-type retainer. They were individually housed in stainless-steel cages (W700  D700  H700 mm) under controlled environmental conditions, including a temperature of 2373 1C, a relative humidity of 55715%, a ventilation rate of 13–16 air exchanges/h, and a 12-h light/12-h dark cycle. The dogs were provided with approximately 300 g of canine standard diet daily (DS-A, Oriental Yeast Co., Ltd., Tokyo, Japan) and allowed free access to tap water, except during the experimental period. This study was approved by our Institutional Committee on the Guide for the Care and Use of Laboratory Animals, and all procedures were performed in accordance with institutional guidelines.

Infusion Solutions containing 33, 67, and 133 mmol/l of citrate were infused via a polytetrafluoroethylene catheter (22G, 0.85 mm outer diameter) placed in the cephalic vein of 4 conscious dogs at 10 ml/kg/h for 4, 2, and 1 h, respectively (corresponding to citrate infusion rates of 0.33, 0.67, and 1.33 mmol/ kg/h, respectively) (Table 1). In addition, as a control group for the 0.33-mmol/kg/h citrate group, 4 conscious dogs were infused with saline solution at 10 ml/kg/h for 4 h.

Clinical signs and electrocardiography Clinical signs were continuously observed up to 2 h after the end of infusion. The time points for recording the ECG in each group are shown in Fig. 1. The heart rate, PQ interval, QRS complex, T wave height, and QT interval in standard limb lead II were evaluated using an animal autoanalyzing electrocardiograph (Fukuda Medical Electronica Co., Ltd., Tokyo, Japan) in conscious dogs restrained in a hangingtype retainer during and after infusion. The corrected QT interval (QTc) was calculated by the logarithmic formula for heart rate in dogs.14 In the saline group, only the heart rate, QT interval, and QTc were measured or calculated. The recording time points for the saline group were the same as for the 0.33-mmol/kg/h citrate group. The ECG was observed up to 2 h after the end of infusion.

ARTICLE IN PRESS 986

T. Fukuda et al.

Table 1

Study design.

Test solution

Infusion rate (ml/kg/h)

Dose of citrate (mmol/kg)

Infusion rate of citrate (mmol/kg/h)

Infusion time (h)

No. of animals

Saline 33 mmol/l Citrate 67 mmol/l Citrate 133 mmol/l Citrate

10 10 10 10

— 1.33 1.33 1.33

— 0.33 0.67 1.33

4 4 2 1

4 4 4 4

 Control for the 0.33-mmol/kg/h citrate group.

• 0.33-mmol/kg/h citrate group Post-infusion (2 h)

Infusion (4 h)

Pre

2

1

3

4

6

• 0.67-mmol/kg/h citrate group Post-infusion (2 h)

Infusion (2 h)

Pre

0.5

1

1.5

2

3

4

• 1.33-mmol/kg/h citrate group Post-infusion (2 h)

Infusion (1 h)

Pre

0.5

1

1.5

2

3

Figure 1 Time points (arrows) for ECG recording and blood sample collection in each citrate infusion group. The time points for the saline group were the same as those for the 0.33-mmol/kg/h citrate group.

Blood tests The time points for collecting blood samples in each group are also shown in Fig. 1. Blood samples were obtained from the cephalic vein on the side opposite the infusion site. The blood was collected into tubes containing a serum-separating agent and centrifuged at 3000 rpm for 10 min at 4 1C after standing for 30–45 min at room temperature. Serum citrate, total Ca, and total Mg levels were measured using an Automatic Analyzer 7170 (Hitachi Ltd., Tokyo, Japan). Serum Ca2+ and Mg2+ levels were measured using a NOVA CRT8 (NOVA Biomedical, Waltham, MA, USA). In the saline group, only serum citrate and Ca2+ levels were measured. The blood collection time points for the saline group

were the same as for the 0.33-mmol/kg/h citrate group.

Statistical analysis All data are expressed as mean7standard deviation (SD), and significant differences from pre-infusion values in the 0.33-, 0.67-, and 1.33-mmol/kg/h citrate groups were assessed by the paired t-test. In addition, mean changes from pre-infusion values of the QT interval, QTc, serum citrate level, and serum Ca2+ level in the 0.33-mmol/kg/h citrate group were compared with those in the saline group, and significant differences were assessed by the Student’s t-test. A P value less than 0.05 was considered statistically significant.

ARTICLE IN PRESS Tolerable infusion rate of citrate

987

Table 2 Summary of clinical signs observed in male dogs during and after citrate infusion at 0.33, 0.67, and 1.33 mmol/kg/h. Citrate group

Findings

During infusion

Post-infusion

Within 1 h Within 2 h Within 4 h Within 1 h Within 2 h 0.33 mmol/kg/h

No. of animals participating No. of animals with no clinical signs

4 4

4 4

4 4

4 4

4 4

0.67 mmol/kg/h

No. of animals participating No. of animals with no clinical signs Nausea Vomiting

4 4

4 3

— —

4 4

4 4

0 0

1 1

— —

0 0

0 0

No. of animals participating No. of animals with no clinical signs Hyperpnea Nausea Vomiting Reddening of the pinna Swelling of the face Salivation Nasal discharge Sneezing

4 0

— —

— —

4 1

4 4

2 4 4 1 1 4 1 1

— — — — — — — —

— — — — — — — —

0 0 0 1 0 3 0 0

0 0 0 0 0 0 0 0

1.33 mmol/kg/h

— No infusion.

Results Effects of citrate infusion on clinical signs The results for clinical signs are summarized in Table 2. No significant changes were observed during or after the infusion of citrate at 0.33 mmol/kg/h. Although no changes in clinical signs were noted up to 1 h after the start of infusion at 0.67 mmol/kg/h, nausea and vomiting were observed in 1 of the 4 animals by 2 h after the start of infusion. However, these signs resolved by 1 h after the end of infusion. At 1.33 mmol/kg/h, hyperpnea, nausea, vomiting, reddening of the pinna, swelling of the face, salivation, nasal discharge, and sneezing were observed during infusion. Hyperpnea, nausea, vomiting, swelling of the face, nasal discharge, and sneezing resolved by 1 h after the end of infusion, and reddening of the pinna and salivation resolved by 2 h after the end of infusion.

Effects of citrate infusion on ECG parameters The results for ECG parameters are shown in Fig. 2. No significant changes were observed in heart rate during the infusion of citrate at 0.33 mmol/kg/h.

However, at 0.67 and 1.33 mmol/kg/h, a significant increase in heart rate was observed during infusion, and this increase was directly related to the infusion rate. The heart rate significantly increased to 138.373.7 beats per minute (bpm) (pre-infusion value ¼ 106.077.4 bpm) and 165.0 716.6 bpm (pre-infusion value ¼ 111.076.2 bpm) after the end of infusion at 0.67 and 1.33 mmol/kg/h, respectively. The heart rate returned to the preinfusion value by 1 h after the end of infusion at 0.67 and 1.33 mmol/kg/h. No significant changes in the PQ interval or QRS complex were observed at any infusion rate. Significant prolongations of the QT interval and QTc were observed during infusion, and these changes were directly related to the infusion rate. When citrate was infused at 0.33 mmol/kg/h, the QTc was significantly prolonged from a pre-infusion value of 193.575.8 ms to a peak value of 214.079.9 ms at 3 h after the start of infusion. At 0.67 and 1.33 mmol/kg/h, the QTc was significantly prolonged during infusion, reaching 225.878.7 ms (pre-infusion value ¼ 190.8710.9 ms) and 238.375.9 ms (pre-infusion value ¼ 188.376.3 ms) after the end of infusion, respectively. The QTc immediately recovered or showed a tendency toward recovery to the preinfusion value at 2 h after the end of infusion at all infusion rates. Although no significant changes in T

(a)

(bpm)

Pre

Pre

1

1

2

2

4

3 Time (h)

4

1.33 mmol/kg/h ×1-h infusion

0.67 mmol/kg/h ×2-h infusion

0.33 mmol/kg/h ×4-h infusion

QTc

3 Time (h)

1.33 mmol/kg/h ×1-h infusion

0.67 mmol/kg/h ×2-h infusion

0.33 mmol/kg/h ×4-h infusion

6

6

(b)

(d)

-14

-12

-10

-8

-6

-4

-2

0

2

4

140

160

180

200

220

240

260

Pre

Pre

1

1

2

3 Time (h)

4

3 Time (h)

4

1.33 mmol/kg/h ×1-h infusion

0.67 mmol/kg/h ×2-h infusion

0.33 mmol/kg/h ×4-h infusion

T wave height

2

1.33 mmol/kg/h ×1-h infusion

0.67 mmol/kg/h ×2-h infusion

0.33 mmol/kg/h ×4-h infusion

QT interval

6

6

Figure 2 Time courses of changes in electrocardiographic parameters during and after citrate infusion. Values are mean7SD for 4 dogs. Significantly different from the pre-infusion value, yPo0.05 and yyPo0.01 in the 0.33-mmol/kg/h citrate group; ]Po0.05, ]]Po0.01, and ]]]Po0.001 in the 0.67-mmol/kg/h citrate group; and *Po0.05, ** Po0.01, and ***Po0.001 in the 1.33-mmol/kg/h citrate group (paired t-test).

140

160

180

200

220

240

260

40

60

80

100

120

140

160

180

Heart rate

988

(c)

(ms)

(ms) (× 0.1 mV)

200

ARTICLE IN PRESS T. Fukuda et al.

ARTICLE IN PRESS Tolerable infusion rate of citrate

989 in citrate or Ca2+ levels were observed from 1 h after the start of infusion to the end of infusion. The ranges for citrate and Ca2+ levels were 1.0770.06 to 1.2270.11 mmol/l (pre-infusion value: 0.3870.01 mmol/l) and 1.2870.03 to 1.3270.06 mmol/l (pre-infusion value: 1.507 0.03 mmol/l), respectively, from 1 h after the start of infusion to the end of infusion. At 0.67 and 1.33 mmol/kg/h, the citrate level significantly increased to 2.0370.12 mmol/l (pre-infusion value: 0.4070.04 mmol/l) and 3.4070.51 mmol/l (pre-infusion value: 0.3370.01 mmol/l), respectively, at the end of infusion. In contrast with the citrate level, the Ca2+ level significantly decreased to 0.9170.03 mmol/l (pre-infusion value: 1.517

wave height were observed during or after infusion at 0.33 or 0.67 mmol/kg/h, T wave height decreased during infusion at 1.33 mmol/kg/h, and a significant change was observed at 0.5 h after the end of infusion.

Effects of citrate infusion on serum citrate, Ca2+, and Mg2+ levels The results for serum citrate, Ca2+, and Mg2+ levels are shown in Fig. 3. Although the mean citrate level increased during infusion in accordance with the infusion rate, the Ca2+ level decreased. When citrate was infused at 0.33 mmol/kg/h, no changes

Serum Ca2+

Serum citrate 1.6

4 0.33 mmol/kg/h ×4-h infusion 0.67 mmol/kg/h ×2-h infusion 1.33 mmol/kg/h ×1-h infusion

1.4 1.2

3

(mmol/L)

(mmol/L)

1 2

0.8 0.6 0.33 mmol/kg/h ×4-h infusion 0.67 mmol/kg/h ×2-h infusion 1.33 mmol/kg/h ×1-h infusion

0.4

1

0.2 0

0 Pre

(a)

1

2

3 Time (h)

4

6

Pre

1

2

(b)

3 Time (h)

4

6

Serum Mg2+ 0.7 0.33 mmol/kg/h ×4-h infusion 0.67 mmol/kg/h ×2-h infusion 1.33 mmol/kg/h ×1-h infusion

0.6

(mmol/L)

0.5 0.4 0.3 0.2 0.1 0 Pre

(c)

1

2

3

4

6

Time (h)

Figure 3 Time courses of serum citrate, Ca2+, and Mg2+ levels during and after citrate infusion. Significantly different from the pre-infusion value, yyPo0.01 and yyyPo0.001 in the 0.33-mmol/kg/h citrate group; ]Po0.05 and ]]]Po0.001 in the 0.67-mmol/kg/h citrate group; and *Po0.05, **Po0.01, and ***Po0.001 in the 1.33-mmol/kg/h citrate group (paired t-test).

ARTICLE IN PRESS 990

T. Fukuda et al.

0.02 mmol/l) and 0.6170.08 mmol/l (pre-infusion value: 1.4870.02 mmol/l) at 0.67 and 1.33 mmol/ kg/h, respectively. The citrate and Ca2+ levels immediately recovered or showed a tendency toward recovery to pre-infusion values by 2 h after the end of infusion at all infusion rates. Compared with the pre-infusion value, the mean Mg2+ level significantly decreased during infusion in accordance with the infusion rate. Although no changes in the citrate level were observed from 1 h after the start of infusion to the end of infusion at 0.33 mmol/kg/h, the Mg2+ level significantly decreased during this period, falling to 0.327 0.03 mmol/l (pre-infusion value: 0.537 0.02 mmol/l) at the end of infusion. When citrate was infused at 0.67 and 1.33 mmol/kg/h, the Mg2+ level significantly decreased to 0.2070.01 mmol/l (pre-infusion value ¼ 0.4970.03 mmol/l) and 0.1770.02 mmol/l (pre-infusion value ¼ 0.557 0.06 mmol/l), respectively, at the end of infusion. The Mg2+ level immediately showed a tendency toward recovery to the pre-infusion value by 2 h after the end of infusion at all infusion rates, but recovery was slower than for the Ca2+ level.

compared with the pre-infusion value during and after infusion at 1.33 mmol/kg/h. On the other hand, the mean total Mg level significantly decreased compared with the preinfusion value during infusion at all infusion rates. The total Mg level did not recover or show a tendency toward recovery to the pre-infusion value by 2 h after the end of infusion at any infusion rate.

Effects of citrate infusion on serum total Ca and total Mg levels

Discussion

Comparison of mean changes in QT interval, QTc, serum citrate level, and Ca2+ level: saline vs. 0.33–mmol/kg/h citrate group The results are shown in Fig. 5. Although prolongations of the QT interval and QTc were observed during saline infusion, serum citrate and Ca2+ levels were not affected. No significant differences in the QT interval or QTc were observed between the saline group and the 0.33-mmol/kg/h citrate group, but during infusion of citrate at 0.33 mmol/kg/h, a significant increase in the serum citrate level and a decrease in the serum Ca2+ level were observed compared with the saline group.

The maximum infusion rate of citrate employed in this study was 1.33 mmol/kg/h because it was reported that hemodynamic changes and prolongation of the QTc were observed at this infusion rate of citrate in patients.15 In addition, we anticipated that the severity of citrate toxicity would be

The results for serum total Ca and total Mg levels are shown in Fig. 4. No significant changes in the mean total Ca level were observed during or after infusion at 0.33 or 0.67 mmol/kg/h. However, the total Ca level significantly increased and decreased

Serum total Mg

Serum total Ca 12

2.2 0.33 mmol/kg/h ×4-h infusion 0.67 mmol/kg/h ×2-h infusion 1.33 mmol/kg/h ×1-h infusion

2

11

1.8 (mg/dL)

(mg/dL)

10

9

1.4

0.33 mmol/kg/h ×4-h infusion

8

1.2

0.67 mmol/kg/h ×2-h infusion 1.33 mmol/kg/h ×1-h infusion

7

1

6

0.8 Pre

(a)

1.6

1

2

3 Time (h)

4

6

Pre

(b)

1

2

3 Time (h)

4

6

Figure 4 Time courses of serum total Ca and total Mg levels during and after citrate infusion. Values are mean7SD for 4 dogs. Significantly different from the pre-infusion value, yyPo0.01 and yyyPo0.001 in the 0.33-mmol/kg/h citrate group; ] Po0.05 and ]]]Po0.001 in the 0.67-mmol/kg/h citrate group; and *Po0.05, **Po0.01, and ***Po0.001 in the 1.33mmol/kg/h citrate group (paired t-test).

(a)

(ms)

-0.2

0

0.2

0.4

0.6

0.8

1

-30

-20

-10

0

10

20

30

40

50

Pre

Pre

+++ +++

Serum citrate

3 Time (h)

1

4

2

3 Time (h)

4-h infusion 4

Saline 0.33 mmol/kg/h of citrate

2

+++ p< 0.001 vs Saline

+++

1

4-h infusion

Not significant vs Saline

0.33mmol/kg/h of citrate

Saline

+++

6

6

(b)

(d)

-0.6

-0.4

-0.2

0

0.2

-30

-20

-10

0

10

20

30

40

50

2

+++ ++ p< 0.01 vs Saline

+++

Pre

1

2

3 Time (h)

4-h infusion

+++

Serum Ca2+

3 Time (h)

Saline 0.33 mmol/kg/h of citrate

1

+++ p< 0.001 vs Saline

Pre

4-h infusion

Not significant vs Saline

Saline 0.33 mmol/kg/h of citrate

QTc

4

4

++

6

6

Tolerable infusion rate of citrate

Figure 5 Mean (SD) changes from pre-infusion values in QT interval, QTc, serum citrate level, and serum Ca2+ level during and after citrate infusion at 0.33 mmol/kg/h. Values are mean7SD for 4 dogs. The infusion rate of saline solution was 10 ml/kg/h, the same as the infusion rate in the 0.33-mmol/kg/h citrate group. Significantly different from saline by the Student’s t-test, ++Po0.01 and +++Po0.001.

(c)

(mmol/L)

(ms) (mmol/L)

QT interval

ARTICLE IN PRESS 991

ARTICLE IN PRESS 992

T. Fukuda et al.

directly related to the infusion rate of citrate because several investigators have reported that the decrease in Ca2+ level is associated with the transfusion rate and the chelating action of citrate in the blood preservative.6–8 In the present study, citrate solutions were therefore infused via the cephalic vein of 4 conscious dogs at citrate infusion rates of 0.33, 0.67, and 1.33 mmol/kg/h up to 1.33 mmol/kg. Because toxic effects of citrate based on clinical signs and the ECG were confirmed when an increase in the serum citrate level and a decrease in the serum Ca2+ level were observed in accordance with the infusion rate at 0.67 and 1.33 mmol/kg/h, they were considered to be related to the decrease in the serum Ca2+ level due to the chelating action of citrate. Many investigators have shown that these toxic effects are also observed in humans or dogs with ionized hypocalcemia resulting from citrate infusion.1,4–6,16–20 However, the neuromuscular signs such as tetany, carpopedal spasm, and seizures that may occur in patients with ionized hypocalcemia 1,7,21 were not observed in the present study. Although no changes in clinical signs were observed during citrate infusion at 0.33 mmol/kg/h, significant prolongation of the QTc was seen compared with the pre-infusion value, when an increase in the serum citrate level and a decrease in the serum Ca2+ level were noted. However, comparable QTc prolongation was also observed during the infusion of saline solution, which was the control for citrate at 0.33 mmol/kg/h, although serum citrate and Ca2+ levels were not affected. It is therefore considered that the prolongation of the QTc observed for citrate at 0.33 mmol/kg/h was not related to a decrease in the serum Ca2+ level but to the actual infusion experiment. Thus, the tolerable infusion rate of citrate in conscious dogs based on clinical signs and the ECG is considered to be 0.33 mmol/kg/h (64 mg/ kg/h), which also induces no significant clinical signs or symptoms in healthy adults.19 In addition, with regard to the commercial parenteral solutions shown in Table 3, the infusion rate of citrate shows a

Table 3

sufficient safety margin, suggesting that there is no risk of changes in clinical signs or the ECG if they are infused at standard infusion rates. Serum citrate and Ca2+ levels reached a steady state from 1 h after the start of infusion to the end of infusion at 0.33 mmol/kg/h of citrate, with a citrate level of about 1.07–1.22 mmol/l (2.8–3.2 times the pre-infusion value) and a Ca2+ level of about 1.28–1.32 mmol/l (0.85–0.88 times the preinfusion value). Therefore, the tolerable peak serum citrate and Ca2+ levels in conscious dogs based on clinical signs and the ECG are estimated to be about 1.22 and 1.28 mmol/l, respectively. However, care is needed when these tolerable serum levels are employed in clinical practice because there are differences between dogs and humans in the normal serum citrate and Ca2+ levels. The normal (pre-infusion) serum citrate and Ca2+ levels in the healthy dogs used in this study were about 0.33–0.45 and 1.45–1.54 mmol/l, respectively. Compared with these values, the reference normal serum citrate and Ca2+ levels in healthy adults are about 0.03–0.23 and 0.98–1.20 mmol/l,19 respectively. Thus, tolerable serum citrate and Ca2+ levels should be considered not in terms of the absolute values but in terms of the relative changes in these values in clinical practice. In fact, it has been reported that the relative changes in serum citrate and Ca2+ levels that were observed at 0.33 mmol/kg/h of citrate do not affect clinical signs, the ECG, or hemodynamic function in clinical practice.19,22 Although the serum citrate and Ca2+ levels reached a steady state at 0.33 mmol/kg/h of citrate, the Mg2+ level fell. Furthermore, the recovery of the Mg2+ level was slower than for the Ca2+ level. The reason for this result is thought to be related to the differences in Ca2+ and Mg2+ homeostasis. Specifically, the Ca2+ level immediately improved with the rapid mobilization of Ca2+ from skeletal stores and the hepatic clearance of citrate,23 but the Mg2+ level did not show immediate recovery because mobilization of Mg2+ out of

Examples of clinical infusion rates of citrate for parenteral solution products.

Parenteral solution product

Clinical infusion rate (ml/kg/h)

Citrate level (mmol/l)

Infusion rate of citrate (mmol/kg/h)

Acetate Ringer’s solution with 1% carbohydrate* 10.5% carbohydrate solution with electrolytesy

15

2

0.03

4.7

0.02

5

The trade names of these products are Physios 140 (*) and TRIFLUIDs (y). All infusion rates were calculated assuming a body weight of 50 kg.

ARTICLE IN PRESS Tolerable infusion rate of citrate the cell takes several days. Although neuromuscular symptoms, as well as ventricular fibrillation and torsade de pointes, which may occur in patients with ionized hypomagnesemia,11,12 were not observed at 0.67 or 1.33 mmol/kg/h of citrate, QTc prolongation may also have been induced by the decrease in the Mg2+ level. Therefore, it is considered that future studies are needed to determine whether the observed QTc prolongation is induced by ionized hypocalcemia or ionized hypomagnesemia resulting from citrate infusion. Although no significant changes in total Ca were observed during or after citrate infusion at 0.33 or 0.67 mmol/kg/h, the total Mg level fell significantly at all infusion rates. These differences in the changes in total Ca and total Mg levels may also reflect the differences in Ca2+ and Mg2+ homeostasis, suggesting that the Ca2+ bound to the infused citrate was rapidly mobilized, but that this was not the case for Mg2+. Incidentally, it should also be considered that the ionized fraction is more important than the complexed fraction of calcium and magnesium circulating in the extracellular fluid, because only the ionized fraction is active and biologically significant. In conclusion, nausea, vomiting, reddening of the pinna, an increase in heart rate (tachycardia), QTc prolongation, etc. were observed when citrate was infused at 0.67 or 1.33 mmol/kg/h. The tolerable infusion rate of citrate in conscious dogs based on clinical signs and the ECG is therefore determined to be 0.33 mmol/kg/h, at which the tolerable serum citrate and Ca2+ levels are estimated to be about 1.22 and 1.28 mmol/l, respectively.

Acknowledgments The authors would like to thank Mr. Kunihiro Yamashita, Mr. Masanobu Hayashi, Mrs. Keiko Ishikawa, and Mrs. Sanae Kishimoto for their expert technical assistance.

References 1. Dzik WH, Kirkley SA. Citrate toxicity during massive blood transfusion. Transfus Med Rev 1988;2:76–94. 2. Scheidegger D, Drop LJ, Schellenberg JC. Role of the systemic vasculature in the hemodynamic response to changes in plasma ionized calcium. Arch Surg 1980;115: 206–11. 3. Corbascio AN, Smith NT. Hemodynamic effects of experimental hypercitremia. Anesthesiology 1967;28:510–6.

993 4. Nakasone N, Watkins Jr E, Janeway CA, Gross RE. Experimental studies of circulatory derangement following the massive transfusion of citrated blood. J Lab Clin Med 1954;43:184–95. 5. Davis TM, Singh B, Choo KE, Ibrahim J, Spencer JL, St John A. Dynamic assessment of the electrocardiographic QT interval during citrate infusion in healthy volunteers. Br Heart J 1995;73:523–6. 6. Kahn RC, Jascott D, Carlon GC, Schweizer O, Howland WS, Goldiner PL. Massive blood replacement: correlation of ionized calcium, citrate, and hydrogen ion concentration. Anesth Analg 1979;58:274–8. 7. Kost GJ, Jammal MA, Ward RE, Safwat AM. Monitoring of ionized calcium during human hepatic function. Critical values and their relevance to cardiac and hemodynamic management. Am J Clin Pathol 1986;86:61–70. 8. Diaz J, Acosta F, Parrilla P, et al. Correlation among ionized calcium, citrate, and total calcium levels during hepatic transplantation. Clin Biochem 1995;28:315–7. 9. Killen DA, Grogan II EL, Gower RE, Collins HA. Response of canine plasma-ionized calcium and magnesium to the rapid infusion of acid-citrate-dextrose (ACD) solution. Surgery 1971;70:736–43. 10. McLellan BA, Reid SR, Lane PL. Massive blood transfusion causing hypomagnesemia. Crit Care Med 1984;12:146–7. 11. Weisinger JR, Bellorin-Font E. Magnesium and phosphorus. Lancet 1998;352:391–6. 12. Gambling DR, Birmingham CL, Jenkins LC. Magnesium and the anaesthetist. Can J Anaesth 1988;35:644–54. 13. Gibson II JG, Rees SB, McManus TJ, Scheitlin WA. A citratephosphate-dextrose solution for the preservation of human blood. Am J Clin Pathol 1957;28:569–78. 14. Matsunaga T, Mitsui T, Harada T, Inokuma M, Murano H, Shibutani Y. QT corrected for heart rate and relation between QT and RR intervals in beagle dogs. J Pharmacol Toxicol Methods 1997;38:201–9. 15. Bunker JP, Bendixen HH, Murphy AJ. Hemodynamic effects of intravenously administered sodium citrate. N Engl J Med 1962;266:372–7. 16. Drop LJ, Scheidegger D. Haemodynamic consequences of citrate infusion in the anaesthetized dog: comparison between two citrate solutions and the influence of beta blockade. Br J Anaesth 1979;51:513–21. 17. Szymanski IO. Ionized calcium during plateletpheresis. Transfusion 1978;18:701–8. 18. Kishimoto M, Ohto H, Shikama Y, Kikuta A, Kimijima I, Takenoshita S. Treatment for the decline of ionized calcium levels during peripheral blood progenitor cell harvesting. Transfusion 2002;42:1340–7. 19. Olson PR, Cox C, McCullough J. Laboratory and clinical effects of the infusion of ACD solution during plateletpheresis. Vox Sang 1977;33:79–87. 20. Weinstein R. Hypocalcemic toxicity and atypical reactions in therapeutic plasma exchange. J Clin Apheresis 2001;16: 210–1. 21. Sorell M, Rosen JF. Ionized calcium: serum levels during symptomatic hypocalcemia. J Pediatr 1975;87:67–70. 22. Martin TJ, Kang Y, Robertson KM, Virji MA, Marquez JM. Ionization and hemodynamic effects of calcium chloride and calcium gluconate in the absence of hepatic function. Anesthesiology 1990;73:62–5. 23. Denlinger JK, Nahrwold ML, Gibbs PS, Lecky JH. Hypocalcemia during rapid blood transfusion in anaesthetized man. Br J Anaesth 1976;48:995–1000.