Materials and Methods

Animals—All procedures were reviewed and approved by the Institutional Animal Care and Use Committee of QTest Labs and complied with federal guidelines for the care and use of laboratory animals. Eight purpose-bred young adult mixed-breed dogs (4 males and 4 females) weighing 12.2 to 18.9 kg (26.84 to 41.58 lb) were acclimated for at least 14 days prior to the start of the study. Dogs were individually housed in a temperature-controlled facility with a 12-hour light-dark cycle and provided commercial dry dog food and water ad libitum. A health examination was performed on each dog 7 and 1 days before being studied. The health examination included a physical examination, fecal flotation test, ophthalmologic assessment, auscultation of the heart and lungs, ECG, hemogram, and serum biochemical analysis.

Design—Dogs were randomly assigned to receive 0, 10, 20, or 30 mL of LRS^a/kg/h (0, 4.5, 9.1, or 13.6 mL/lb/h) in a 4-period, 4-treatment crossover study with repeated measures. A block randomization method was used to randomize subjects into groups of equal size for each treatment. All possible balanced treatment combinations within a block were calculated and then blocks were randomly chosen. All 8 dogs received each treatment for 60 minutes during isoflurane anesthesia with a minimum of 7 days between treatments. The study was not conducted in a blinded fashion.

Animal preparation and procedures—Blood and urine samples were collected 24 hours before (day −1), during (day 0), and 24 hours after (day 1) treatment for each experiment. Urine samples were collected aseptically via cystocentesis by use of a 23-gauge needle and a 10-mL syringe (days −1 and 1) and bladder catheterization (day 0). The glomerular filtration rate was determined on the basis of exogenous creatinine clearance on days −1 and 1.¹⁹

On the day that the dog was assigned to receive LRS (0, 10, 20, or 30 mL/kg/h), a 20-gauge IV catheter was inserted into the cephalic vein. All dogs received acepromazine (0.025 mg/kg [0.0114 mg/lb], IV) and butorphanol (0.1 mg/kg [0.05 mg/lb], IV) 10 minutes before receiving propofol (approx 4 to 6 mg/kg [1.82 to 2.7 mg/lb], IV, to effect). The dogs were orotracheally intubated and mechanically ventilated by use of an ascending-bellows, volume-cycled, pressure-regulated ventilator.^b The ventilator was set to deliver a tidal volume of 10 to 15 mL/kg (maximum allowed pressure, 16 to 20 cm H₂O) at a rate of 8 to 12 breaths/min to maintain end-tidal partial pressure of carbon dioxide^c between 35 and 45 mm Hg. The endotracheal tube was connected to a circle anesthetic rebreathing circuit, and anesthesia was maintained with isoflurane in oxygen delivered by use of an out-of-circuit precision vaporizer.^d The end-tidal inhalant concentration^c was maintained between 1.4% and 1.6%. A warm air blanket^e was used to maintain body temperature between 36.5° and 37°C (97.7° and 98.6°F). The ECG, hemoglobin oxygen saturation, and arterial blood pressure (oscillometric technique^f) were continuously monitored throughout induction and maintenance of isoflurane anesthesia.

A flow-directed, balloon-tipped 5F thermodilution catheter^g was percutaneously and aseptically inserted into the jugular vein and advanced into the pulmonary artery to permit continuous monitoring of central venous and pulmonary artery pressure and intermittent determination of cardiac output by use of thermodilution.^h The mean of 3 thermodilution cardiac output determinations was calculated to obtain a cardiac output value at each recording time point. A catheter-tipped pressure-sensing catheter with a fluid-filled sampling portⁱ was aseptically inserted into the right femoral artery and the tip advanced into the thoracic aorta. This catheter was used to record arterial blood pressure and to collect anerobic blood samples for determination of arterial pH^j and blood gases (Pao₂ and Paco₂).^j Mixed venous blood samples were obtained from the pulmonary artery catheter for determination of mixed venous pH,^j venous blood gases (Pvo₂ and Pvco₂),^j and hematologic and serum biochemical data. Each dog was in a stabilized anesthetic state for approximately 30 minutes before recording baseline data (time 0), collection of blood samples, and initiation of fluid administration (0, 10, 20, and 30 mL/kg/h^k). All dogs were allowed to recover from anesthesia, and the times to tracheal tube removal, sternal recumbency, and standing were recorded.

Experimental data—Urine samples were evaluated for color, clarity, bilirubin, blood, protein, glucose, ketones, urobilinogen, microscopic sediment, nitrites, occult blood, leukocytes, pH, specific gravity, and electrolytes (Na⁺, K⁺, and Cl⁻). The plasma BUN and serum creatinine concentrations were determined on days −1, 0, and 1. The glomerular filtration rate (reference range, 2.5 to 5.5 mL/kg/min [1.14 to 2.5 mL/lb/min]) was assessed on days −1 and 1.²⁰

Blood hemoglobin concentration^j and serum total protein,^l albumin,^m lactate,^j and electrolyte (Na⁺, K⁺, Cl⁻, and ionized Ca²⁺)^j concentrations; COPⁿ; and PCV were determined from mixed venous blood samples at time 0 (baseline) and 30 and 60 minutes. Arterial and mixed venous pH,^j Po₂,^j and Pco₂^j were determined from anaerobic blood samples collected at baseline and at 30 and 60 minutes. Whole blood and plasma viscosities^o were determined from mixed venous blood samples obtained at baseline and 60 minutes at a shear rate of 150 s⁻¹ and temperature of 37°C. Heart rate^p; respiration rate; systolic, mean, and diastolic arterial blood pressures^p; thermodilution cardiac output (0 and 60 minutes only); central venous^p and mean pulmonary artery pressure^p; and body temperature^p were recorded at 0 (baseline), 15, 30, and 60 minutes. Systemic and pulmonary vascular resistance, stroke volume, pulse pressure, and oxygen delivery were calculated at 0 and 60 minutes from standard formulas. The bladder was emptied, and urine output was recorded from time 0 for 60 minutes.

Blood volume and PV changes over time were calculated from estimated baseline values of blood volume (81.5 mL/kg [37.05 mL/lb]) and PV (49 mL/kg [22.3 mL/lb]) for clinically normal dogs.²¹ Preinfusion and subsequent estimates of blood volume and PV were determined by use of previously described mass balance equations.^22,23 Preinfusion calculation of blood volume (PV_t/[1 – PCV × 0.95 × 0.92]) was corrected for known errors in large vessel Hct (0.92) and plasma trapped in erythrocyte column of the Hct tube (0.95).²⁴ Changes in PV expansion were calculated as ΔPV = PV_t – PV_t−1. Volume expansion efficiency was calculated as the net change in PV divided by the cumulative volume of LRS infused (ΣF_t) such that VEE_t = PV/ΣF_t.²⁴

Statistical analysis—An ANOVA appropriate for a 4-period, 4-treatment crossover study with repeated measures was used to evaluate the effect of fluid administration on the outcomes of interest. The assumption of normality was evaluated through the use of normal probability plots. The statistical model included treatment sequence, period, treatment (0, 10, 20, or 30 mL/kg/h), time, and the time by treatment interaction as fixed effects. Dog nested in sequence was included as a random effect. Given only 2 to 4 generally equally spaced time intervals, the structure of the covariance matrix was assumed to be compound symmetric. The repeated measures associated with period were ignored. If the time by treatment interaction was significant, an all-pairwise comparison of treatment effects within time was conducted. Additionally, given a time by treatment interaction, within treatment, time effects were assessed by comparing time 0 values to values collected after time 0. If no interaction existed, the main effect of time was investigated and again an all-pairwise approach was used. Where there were no repeated measures, the same approach was taken but the time and time by treatment interactions were excluded from the model. Values of P < 0.05 were considered significant.

Results

Urine pH ranged from 5.0 to 9.0 and urine specific gravity from 1.018 to 1.042 on days −1, 0, and 1. The plasma BUN concentration ranged from 13 to 35 mg/dL and serum creatinine concentration from 0.5 to 1.2 mg/dL on days −1, 0, and 1. The glomerular filtration rate ranged from 3.0 to 5.4 mL/kg/min (1.4 to 2.45 mL/lb/min) and was within reference limits for dogs on days −1 and 1.

Induction of anesthesia and transition to isoflurane anesthesia were uneventful. No dog became excited or had uncharacteristic changes in heart rate or rhythm, oxygen saturation (< 90%), or oscillometric blood pressure measurements (systolic arterial blood pressure < 90 mm Hg). The normal probability plots did not identify deviations from normal, and thus, all outcomes were analyzed under the assumption of normality. The IV administration of LRS produced dose-dependent decreases in PCV; blood hemoglobin concentration and serum total protein and albumin concentrations; COP; and whole blood viscosity, compared with baseline values (Table 1). Plasma viscosity did not change. These responses were greater for administration at 30 mL/kg/h, compared with administration at 20 and 10 mL/kg/h and baseline. Serum Na⁺ (140 to 145 mEq/L), K⁺ (3.3 to 4.8 mEq/L), Cl⁻ (109 to 116 mEq/L), and ionized Ca²⁺ (1.36 to 1.51 mmol/L) concentrations did not change during fluid administration and were not different among treatment groups. Arterial and mixed venous pH, blood gas variables (Po₂, and Pco₂), and lactate concentrations did not change during LRS administration (Table 2). Increasing rates of LRS administration did not change heart rate (Table 3). Systemic arterial pressures increased minimally but not significantly during LRS administration and were not different among treatment groups. However, mean central venous pressure apparently increased with LRS administration but not significantly (P = 0.089). Increasing rates of LRS administration increased pulmonary artery pressure and CO. The increases in pulmonary artery pressure were greatest during LRS administration at 30 mL/kg/h, followed by administration at 20 mL/kg/h, and were not observed at baseline or during LRS administration at 10 mL/kg/h. The increases in cardiac output were greatest during LRS administration at 30 mL/kg/h, followed by LRS administration at 20 and 10 mL/kg/h, and were not observed at baseline. The rate of urine production did not change and was not different among treatment groups (0.7 ± 0.4 mL/kg/h to 1.1 ± 0.8 mL/kg/h [0.32 ± 0.18 mL/lb/h to 0.5 ± 0.36 mL/lb/h]). Stroke volume increased and systemic vascular resistance decreased during LRS administration (Table 4). The increases in stroke volume and decreases in systemic vascular resistance were greatest during administration at 30 mL/kg/h, followed by administration at 20 mL/kg/h, and were not observed at baseline or during LRS administration at 10 mL/kg/h. Pulmonary vascular resistance and pulse pressure were not changed by LRS administration. Body temperature ranged from 35.1° to 37.9°C (95.18° to 100.22°F) and did not change with LRS administration. Increasing rates of LRS administration increased PV (Figure 1). The 60-minute increase in PV was greatest during LRS administration at 30 mL/kg/h, followed by LRS administration at 20 mL/kg/h, and was not significant for baseline measurements and LRS administration at 10 mL/kg/h. The VEE (vol/vol of LRS infused) ranged from 37% to 50% for the 20 and 30 mL/kg/h infusion rates immediately at the end of infusion.

Mean ± SD change in PV produced by LRS infusion to isoflurane-anesthetized dogs. *Values within a treatment group are significantly (P < 0.05) different from time 0. ^a–cValues within a time point with no common superscripts are significantly (P < 0.05) different from each other (n = 8 dogs in a 4-treatment cross-over study).

Citation: Journal of the American Veterinary Medical Association 239, 5; 10.2460/javma.239.5.630

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SD change in PV produced by LRS infusion to isoflurane-anesthetized dogs. *Values within a treatment group are significantly (P < 0.05) different from time 0. ^a–cValues within a time point with no common superscripts are significantly (P < 0.05) different from each other (n = 8 dogs in a 4-treatment cross-over study).

Citation: Journal of the American Veterinary Medical Association 239, 5; 10.2460/javma.239.5.630

• Download Figure
• Download figure as PowerPoint slide

Mean ± SD change in PV produced by LRS infusion to isoflurane-anesthetized dogs. *Values within a treatment group are significantly (P < 0.05) different from time 0. ^a–cValues within a time point with no common superscripts are significantly (P < 0.05) different from each other (n = 8 dogs in a 4-treatment cross-over study).

Citation: Journal of the American Veterinary Medical Association 239, 5; 10.2460/javma.239.5.630

• Download Figure
• Download figure as PowerPoint slide

Mean times to extubation (13 to 19 minutes), sternal recumbency (26 to 36 minutes), and standing (39 to 49 minutes) were not different among treatment groups. All dogs recovered from anesthesia uneventfully and were returned to their kennels.

Discussion

We investigated the effects of increasing infusion rates of LRS on hematologic, serum biochemical, rheological, hemodynamic, and renal variables in isoflurane-anesthetized dogs. Our results support and extend those of previous studies^{16–18,25,26} in anesthetized dogs, pigs, sheep, and humans and confirm that the administration of an acellular, albumin-free, minimally hypotonic (273 mOsm/L) crystalloid solution (LRS) to isoflurane-anesthetized dogs decreases PCV; serum total protein and albumin concentrations and blood hemoglobin concentration; COP; and whole blood viscosity. Heart rate, mean central venous pressure, and systemic arterial blood pressure did not change. However, as a result of an increase in stroke volume, cardiac output increased from that measured at either baseline or after the infusion of LRS at 10 mL/kg/h; it increased significantly after infusions of 20 or 30 mL/kg/h. The increase in stroke volume may be attributed to increased preload, positive cardiac contractile effects, decreases in aortic input impedance, or a decrease in blood viscosity; however, considering it followed fluid replacement, changes in blood viscosity or preload appear to be the most reasonable explanations.^25,26 The pulmonary artery pressure increased at the highest fluid infusion rate, suggesting compensatory rheological effects and intravascular volume expansion. The rate of urine production did not change among treatment groups during isoflurane anesthesia despite an increasing rate of fluid infusion, suggesting fluid retention during anesthesia and an anesthesia-associated decrease in urine production.

The results of the present study agree with those of other studies^10,25–28 that found that mean central venous and systemic arterial blood pressures are relatively poor monitors of fluid administration when LRS is administered at a rate of 10 mL/kg/h. Cardiac output increases at rates of fluid administration exceeding 20 mL/kg/h, dogs retain large amounts of fluid during isoflurane anesthesia, and urine production is not an adequate indicator of fluid balance in isoflurane-anesthetized dogs. A previous study²⁷ investigated the hematologic and hemodynamic effects of rapid infusion of LRS (90, 225, and 360 mL/kg/h [40.9, 102.3, and 163.6 mL/lb/h]) for 1 hour to mildly dehydrated, conscious dogs. The investigators in that study²⁷ observed that an infusion rate of 90 mL/kg/h was well tolerated, and signs of fluid excess (profuse nasal serous discharge, restlessness, coughing, abdominal distention, vomiting, and diarrhea) were not observed until an infusion rate of 360 mL/kg/h was obtained. These adverse effects resolved quickly after cessation of fluid administration, suggesting that relatively rapid rates of fluid administration (90 mL/kg/h) are tolerated, at least in conscious dogs.²⁷ In agreement with results of the present study, that previous study²⁷ revealed that systemic arterial blood pressure increased minimally and inconsistently and that central venous pressure did not change when LRS was administered at 90 mL/kg/h. More recent studies⁸ investigating the hemodynamic effects of IV infusion of LRS at a rate of 80 mL/kg/h (36.4 mL/lb/h) for 1 hour during isoflurane-induced hypotension (systolic arterial blood pressure, 80 mm Hg) in dogs demonstrated decreases in PCV, COP, whole blood viscosity, and serum total protein and albumin concentrations and no change in arterial blood pressure. Central venous and pulmonary artery pressures, stroke volume, cardiac output, and blood volume increased and systemic vascular resistance decreased 15 minutes after beginning fluid infusion. Urine production was not measured. A clinical study⁵ revealed that the administration of 5, 10, or 15 mL of polyionic crystalloid solution/kg/h, IV, to halothane-anesthetized dogs produced no effect on PCV, serum total protein concentration, or systolic arterial blood pressure. However, subsequent studies¹⁶ in isoflurane- or sevoflurane-anesthetized dogs demonstrated that administration of 10 mL of LRS/kg/h, IV, for 1 hour was associated with a decrease in PCV and blood hemoglobin and serum total protein concentrations. Others have reported similar decreases in PCV, serum total protein concentration, and COP but cautioned that anesthesia alone can decrease serum total protein concentration and COP.¹⁷ A recent clinical study²⁸ conducted in isoflurane-anesthetized dogs administered 10 mg of LRS/kg/h, IV, and subjected to an orthopedic surgical procedure concluded that a large amount of IV fluid is retained as indicated by increased body weight, positive fluid balance, increased total body water volume, and increased extracellular fluid volume. Collectively, the results of our investigation and data provided by others provide evidence that conventional rates of IV administration of LRS dilute blood components and that the fluids are distributed to extravascular fluid depots.

Acid-base abnormalities are common during anesthesia, and the IV administration of crystalloids has the potential to alter acid-base balance. Crystalloids with a strong ion difference of 24 mEq/L maintain base excess near zero and are considered balanced.²⁹ Alternatively, physiologic saline solution has a strong ion difference of 0 mEq/L and the potential to produce hyperchloremic metabolic acidosis regardless of dilution of plasma proteins (weak acids).^29–32 Lactated Ringer's solution is minimally hypotonic and has a strong ion difference of 27 mEq/L, suggesting that it produces minimal effects on strong ion difference and base excess when administered to normovolemic dogs. We did not observe or expect any changes in arterial or venous acid-base status (pH and Pco₂). Our data and those of others^{16–18,29,32,33} confirm that the IV administration of moderate to large amounts of LRS to normovolemic, isoflurane-anesthetized dogs is unlikely to exacerbate acid-base abnormalities.

The rate and duration of fluid administration are key factors in restoring blood volume and hemodynamics. However, higher fluid infusion rates than are currently recommended (> 10 to 15 mL/kg/h) may not have the desired effect upon PV but do increase interstitial fluid volume.^34–37 We estimated the VEE (ΔPV/total cumulative fluid volume) following IV administration of LRS at rates of 20 and 30 mL/kg/h to be approximately 40%, similar to that of a previous study,²⁴ and did not determine the rate of return of vascular volume to baseline. A study²⁴ in isoflurane-anesthetized sheep revealed that the VEE at the end of a 20-minute infusion of saline solution (25 mL/kg [11.4 mL/lb]) was approximately 40% but rapidly decreased to approximately 14% by 30 minutes after infusion. A recent investigation in isoflurane-anesthetized dogs subjected to an orthopedic surgical procedure and administered 10 mL of LRS/kg/h, IV, suggested relatively rapid loss of fluid from the vascular space and retention of large amounts of fluid in the extracellular space.²⁸ Although the experimental methods used in each of the previous studies varied considerably, they emphasize that the rate and duration of fluid administration affect VEE, larger fluid volumes and faster rates of crystalloid administration increase VEE, and crystalloids are not retained within the vascular fluid compartment for extended periods. Furthermore, the increase in VEE may not be associated with an increase in oxygen delivery, regardless of increases in cardiac output, because blood hemoglobin concentration decreased (dilutional effect). Furthermore, our data support the finding that elimination of the infused volume of LRS from the vascular compartment in isoflurane-anesthetized dogs cannot be accounted for by urinary excretion because urine production did not change. This finding supports the contention that most of the infused fluid must accumulate in a peripheral or extravascular site. Additional studies are required to determine the optimal blood hemoglobin concentration relative to cardiac output to maintain an adequate oxygen delivery during LRS infusion in isoflurane-anesthetized dogs.

We attempted to design our study to include and mimic the clinical conditions encountered during isoflurane anesthesia in otherwise healthy dogs. The results of our study must be interpreted carefully. Our study may have been underpowered regarding several hemodynamic variables, including mean central venous and arterial blood pressures, which may have resulted in significant changes had we included more dogs. Our dogs were ventilated, and a specific anesthetic protocol was selected. Anesthesia is known to alter homeostasis and the rate of elimination of infused crystalloids from the intravascular space through changes in their distribution and alterations in urinary excretion.²⁴ None of our dogs were subjected to major surgery or became hypotensive, effects that are known to increase the amount of fluid that accumulates in extravascular body fluid compartment and delays fluid elimination, respectively.^36,37 Our dogs were not splenectomized, which may have altered our calculation of PV and VEE. Furthermore, we may have overestimated the VEE due to alterations in the large-vessel whole-body PCV ratio.^24,36 Finally, we administered a range of fluid rates that encompassed conventional recommendations but may not represent universal clinical practice. Larger infusion rates administered over shorter periods are generally recommended for the acute treatment of hypotension or hemorrhage and likely would have produced different results.

In conclusion, findings in our study indicate that conventional rates of LRS administration to healthy isoflurane-anesthetized dogs increase PV and cardiac output and decrease blood viscosity and the colligative properties of blood but do not increase the rate of urine production or produce consistent and significant alterations in mean central venous or arterial blood pressures.