Experimental critical care in ventilated rats: Effect of hypercapnia on arterial oxygen-carrying capacity

Experimental

Critical Care in Ventilated Rats: Effect of Hypercapnia on Arterial Oxygen-Carrying Capacity

Dan Torbati,

Bala R. Totapally,

Maria

T. Camacho,

and Jack Wolfsdorf

Purpose: We have previously demonstrated an increased arterial O,-carrying capacity in normal ventilated dogs subjected to both acute and prolonged exogenous hypercapnia. In the present study, we tested if arterial hypercapnia, during controlled ventilation, can increase O,-carrying capacity also in rats. Materials and Methods: Twenty young male Sprague Dawley rats were anesthetized (60 mg/kg pentobarbitall, tracheostomized, intubated, and one femoral vein and artery were cannulated. Anesthesia and paralysis were maintained using 15 mg/kg/h pentobarbital intravenously, and 2 mg/kg/h vecuronium bromide. The fluid balance (5 ml/kg/h saline), normothermia, and minute volume were maintained. The mean arterial blood pressure and heart rate were continuously monitored. Experiments included the following: (I) a control group, ventilated with normoxic air for 150 minutes (n = 5); (2) mild hypercapnia, a group of eight rats ventilated with normoxic air for 30 minutes and then ventilated with a mixture of normoxic air at 60 mm Hg COz (8 kPa) for 1 hour; and (3) severe hypercapnia, a group of seven rats were treated exactly as in group II, except a 90 mm Hg (12 kPa) CO, during hypercapnia. Gas-exchange profile, arterial hemoglobin (Hb)

concentration, arterial Hb-oxygen saturation (Hb-O,), and arterial O2 content were periodically determined during normocapnia and 1 hour of hypercapnia. Results: Exposures to mild and severe hypercapnia, in -with maintained ventilation, significantly reduced the arterial O2 content by 20% and 33%, respectively, without significant changes in the arterial Hb concentration (-2%). Severe hypercapnia generated a significant reduction of -14% in the Pao,, but not in P,o,l Fio, ratio. Conclusion: Rats subjected to controlled ventilation and permissive hypercapnia, unlike dogs and perhaps humans, show no augmentation of Hb concentration. Hypercapnia in rats also provokes much stronger Bohr effect than in dogs. Hypercapniainduced Bohr effect in rats is accompanied with extreme desaturations of Hb-02, and substantial reduction in the O,-carrying capacity. We speculate that the strong hypercapniainduced Bohr effect in rats may prevent hypoxia at the tissue level. However, to maintain a stable oxygencarrying capacity in rats used for pulmonary critical care studies with hypercapnia, we suggest to use hyperoxia, with or without a mild hypothermia. Copyright 0 7999 by W.B. Saunders Company

A

of hemoglobin and a relatively weak Bohr effect in dogs.7,8We speculated that hypercapnia-induced augmentation of Hb could be caused by splenic contraction, loss of plasma volume to interstitial fluid, or into the intracellular fluid.* The increased Hb in the blood in response to hypercapnia in dogs is consistant with Hb increases (-16%) found in anesthetized ventilated pigs exposed to extreme hypercapnia (20%), at FiO2of 0.4.9 Indirect evidence also suggests that both normal subjects and emphysematous patients are capable of increasing their hematocrit significantly, on exposure to hypercapnit-normoxia of 3% and 5% C02.7.‘o The “large” mammalian models used for critical care studies, although practically convenient, are also expensive and require special care. The small

MAJOR THERAPEUTIC goal in pulmonary critical care is to reverse the acute effects of severe arterial hypercapnia and hypoxemia.‘z2 However, vigorous ventilatory measures to reduce arterial CO1 built-up; may lead to barotrauma.* Permissive hypercapnia, induced by controlled ventilation, can minimize the need for vigorous ventilatory support,334and thereby may reduce the ventilatorinduced lung injury (VILI). Theretically the effect of increased Pace, during controlled ventilation is expressed by the alveolar gas equation, in which raising the inspired PC02 (Pica?) will be accompanied by reduction in the alveolar PO,, followed by reduction in the Pao,. Additionally, the reduction in the arterial pH during hypercapnia causes a right shift in the Hb-0, dissociation curve that may reduce the arterial O2 content. However, studies suggest that arterial hypercapnia in ventilated animals may actually improve gas-exchange efficiency by reducing dead-space ventilation (bronchoconstriction), increasing the Bohr effect,5 and by increasing collateral ventilation.6 Recently, we have demonstrated that both acute and prolonged arterial hypercapnia can increase the O,-carrying capacity in ventilated dogs and can sustain a normal oxygenation7a8 These reactions were associated with augmentation Journal

of Critical

Care, Vol 14, No 4 (December),

1999: pp 191-197

From the Division of Critical Care Medicine, Miami Children k Hospital, Miami, FL. Received July 28, 1999. Accepted August 9, 1999. Supported by Miami Children k Hospital Foundation k grant to Dan Torbati. Address reprint requests to Dan Torbati, PhD, Division of Critical Care Medicine, Miami Children’s Hospital, 3100 SK 62nd Ave, Miami FL 331.56. Copyright 0 1999 by WB. Saunders Company 08%9441/99/1404-0006$10.00/0 191

TORBATI

192

animal models, with relatively comparable physiology, have the advantage of containing the cost and thus allowing to compare several groups. However, the gas-exchange profiles and buffering capacity are different in normal spontaneously breathing rats, dogs, and humans, subjected to hypercapnia alone or with hypoxia. ‘l-l3 For example, due to a relatively high level of buffering capacity, a smaller increase in [H+], per unit increase in Paoz occurs in rats, compared with humans or dogs [0.62 nmol/ L/mm Hg in rats, compared with 0.77 in humans and dogs12,13].These differences may also occur during pulmonary critical care conditions, simulated in rats. In this study, therefore, we investigated whether rats, likewise dogs, are also capable of increasing their 02-carrying capacity during control ventilation with permissive hypercapnia. MATERIALS

AND METHODS

Anesthesia and Surgery The experimental protocol for this study was approved by the Institutional Animal Care and Use Committee of the Mount Sinai Medical Center, Miami Beach, Florida. The care and handling of the animals used in this study were in accord with National Institute of Health Guidelines for ethical animal research. Twenty young Sprague Dawley (Long Evans strain) male rats (250 to 350 g) were used for this study. Generalized anesthesia was induced with a total of 60 mgkg pentobarbital (given IP, IM, and IV). During surgery, the adequacy of anesthesia was verified by stimulating the withdrawal reflex of the hind limbs. In the supine position, the rats were then tracheostomized and an endotracheal tube (polyethylene tube #240; 3.5-cm long) was placed approximately 1 cm above the carina. A femoral vein and an artery were then cannulated with a 8-to-10 cm polyethylene tube (#lo) that was heat fused to a wider polyethylene tube (#60). Two hours after the induction of anesthesia and completion of all invasive procedures, each rat was placed in a transparent acrylic chamber (18 X 28 X 12 cm; 6-L capacity) in the supine position. The heart rate (HR), mean arterial blood pressure (MABP), and rectal temperature were continuously monitored (Datascope Corp., Model 2001A, Paramus NJ; Physitemp Thermalert, Model TH5; rats thermal probe TX-~, Clifton NJ). To maintain stable anesthesia, the rats were continuously infused with 15 mg/kg/b pentobarbital, IV. The maintenance dose of pentobarbital was dissolved in saline and used also to maintain the fluid balance (5 ml/kg/h). Depth of anesthesia was judged by the relative stability of the HR and MABP, as previously described.‘” To prevent the effects of confounding factors that are usually involved in critical care conditions (ie, lung injury, sepsis, trauma, and oxygen therapy), we purposely used normal rats that were ventilated with regular air.

Ventilator Settings and Experimental Groups The rats’ endotracheal tube was connected to a rodent ventilator (Columbus Instruments, Model CIV-101, Columbus: OH). Initially, the ventilator settings included frequencies of 60 to 80

ET AL

per minute with tidal volumes of 8 to 10 mL/kg. The peak inspiratory pressure varied between 15 and 20 cm H20. The positive end-expiratory pressure was permanently set at 4 cm HzO, with an inspiratory:expiratory ratio of 1:2. In all rats, the artificial ventilation was initially conducted in the closed chamber, continuously circulated with 2 L/min of normoxic air. The airflow into the chamber passed through a controlled heating system, and an ambient temperature of 32°C to 34°C was generated inside the chamber. The inflow port of the ventilator was connected with a tygon tubing that was extended into the chanber. Thus, the warmed air of the chamber was used to ventilate the rats. This arranagement also allowed to produce exact hypercapnic-notmoxic mixtures within a few minutes, The CO2 concentration was continuously monitored by placing the probe of an end tidal COZ monitor near the inflow tube of the ventilator, placed inside the chamber (Novametrix Medical Systems Inc., Model 1260, Willigford, CT). Fifteen minutes after starting the mechanical ventilation, the rats were paralyzed with a bolus injection of 1 mglkg vecuronium bromide IV. The paralysis was maintained with bolus injections of 1 mg/kg vecuronium bromide every 30 minutes (a total of 2 mg/kg/h). The rectal temperature was maintained within 37.5 % 0.3”C. Starting 30 minutes after the initial paralysis, two arterial blood samples (0.25 mL each) were taken for determination of blood gas profiles and co-oximetry (Radiometer, Model ABL-30, and Model OSM3, respectively). All blood samples were corrected for slight changes in the rectal temperature, and the co-oximeter was adjusted for analysis of rab Hb and hemoglobin-oxygen saturation (Hb-OZ). The O2 content was calculated as: (Hb X 1.34) X Hb-O2 + (PaO, X 0.003). To replace the volume of each blood sample, the arterial catheter was washed by 0.25 mL heparin-saline solution, containing 100 U/mL. These pre-experimental blood samples were used to set the final values of frequency, tidal volume, and peak inspiratoty pressure, for the actual isocapnic-normoxic and hypercapnic-normoxic experiments. At this stage, the rats were randomly assigned to the following experimental groups, and the minute volumes were maintained. Group I (controlsj. Five rats in this group were ventilated with normoxic air without CO2 for a period of 150 minutes after the final ventilator settings were established. Every 30 minutes during this period, a blood sample was withdrawn for gasexchange and co-oximetty analyses. The values of each investigated variable at the first 30.minute sampling period was considered as the control for comparison with other samples. At the end of the experiments, the rats were euthanized by a lethal dose of pentobarbital. The thorax was carefully opened and the lungs were isolated and carefully examined for gross signs of pulmonary or cardiac damage. Group II {mild hypercapnic-normoxia). Eight rats were used for this group. Thirty to 60 minutes after the establishment of the final ventilator settings, arterial blood samples were taken and used as normoxic baselines for gas-exchange profiles, Hb concentration, calculation of the OZ content, and HbOZ. Immediately after the second normoxic sample (considered control), a small flow of CO2 was allowed into the chamber. Within 5 to 7 minutes, a Pica? of 60 mm Hg (8 kPa) was established inside the chamber. This value was verified by both a continuous capnography and gas samples that were immediately analyzed by the blood gas analyzer. The gas samples were also used to determine the value of the inspired pressure of oxygen (PioJ

PERMISSIVE

HYPERCAPNIA

IN VENTILATED

RATS

193

for the calculation of the Fioz. Mild hypercapnia in this group lasted for 1 hour during which two blood samples were removed at 30 and 60 minutes. Subsequently, the flow of CO2 into the chamber was terminated and the rats were again ventilated with nonnoxic air for 30 minutes. At the end of this period the last blood sample was taken and rats were euthanized for autopsy. Group III (severe hypercapnic-normoxia). Seven rats in this group were treated exactly as in group II, except receiving 90 mm Hg (12 l-Pa) CO2 during the 60 minutes of hypercapnic ventilation

The reductions were in order of 6% and 8% (P < .05), at periods corresponding to 30 and 60 minutes of hypercapnia in rats ventilated with mild and severe hypercapnia. No significant differences occurred in the arterial Hb-02, and the ratio of PaoJ Fio,, used as an index of gas exchange efficiency throughout the experiment.

Statistical Evaluation

One of the eight rats in this group died on return to normoxic ventilation from hypercapnic challenge. The data analysis in this group was therefore performed with a total of seven rats. The Hb concentration was not significantly changed during mild hypercapnia (0% and 2% at 60- and 90-minute periods), but it was significantly reduced after 2 hours (Table 2). Despite extreme increases in Pace, during mild hypercapnia, no significant changes occurred in the Pao,. However, the arterial Hb-O2 was significantly reduced due to a strong Bohr effect (Fig. l), resulting in a significant reduction in O+arrying capacity during mild hypercapnia (Table 2). However, the significant increase in the arterial [H’] did not affect the Pao,/Fioz ratio, as an index for oxygen diffusion capacity (Fig. 2).

Group II (mild hypercapnic-normoxia)

All values are presented as mean +- SD. Mortality due to hypercapnia, and randomization, resulted in a different sample size for each group. Final statistical analysis was performed with n = 5 rats in the control group; n = 6 rats in the group subjected to mild hypercapnia, and n = 5 in the group exposed to severe hypercapnia. Incomplete data in three nonsurviving rats were evaluated separately. The values of each parameter within the same group were first analyzed by repeated measures of analysis of variance (ANOVA), followed by the Dunnett multiple comparison test. The second baselines (30 minutes in group I; samples before the hypercapnic ventilation in groups II and III were used as the controls. The P values less than 5% were considered significant. Differences among the three groups, for each parameter, were compare with ANOVA followed by Tukey-Kramer multiple comparisons test.

RESULTS

Group I (controls)

Group III (severe hypercapnic-normoxia)

The gas-exchange values during 150 minutes of continuous normoxic ventilation were relatively stable (Table 1). A gradual and significant decrease in Hb concentration and O2 content were observed after removal of several blood samples (Table 1). Table

1. Gas Exchange

and O,-Carrying in Anesthetized,

One of the seven rats in this group died during severe hypercapnia, and another one died after return to normoxia from hypercapnia. The data analysis in this group was therefore performed using five

Capacity Paralyzed,

During

Continuous

and Ventilated

Isocapnic-Normoxia

Rats

Time (minutes) Variables P.&o2 (mm

30

35.5

-t 3.3

97.9

i

35.5

60

90

120

2 4.4

35.3

k 2.7

33.4

2 4.5

i

5.6

98.7

-c 8.9

103.7

t

t

0.05

k 0.02

7.353

+ 0.05

150

34.0

k 3.2

32.8

t

2.1

101.8

k 8.7

103.3

? 7.6

7.337

7.355

-t 0.03

Hg)

p&h (mm

0

6.6

102.5

8.6

Hg)

PH

lHCO,-I,

7.393 21.1

? 0.04 i

7.368

0.7

19.9

-c 1.3

7.353

19.0 t

0.8

t

0.04

18.0 + 0.7*

17.6 t

0.6*

17.8 2 0.4*

(mmol/L) Hb WdL) Hb-0,

13.4 2 0.3

13.5 -t 0.4

12.7 2 0.7*

12.4 i- 0.5*

12.6 -t 0.7*

12.3 i

95.8

95.6

94.4

97.0

95.2 2 3.0

95.1

16.4 2 0.8"

15.9 i

0.9*

496 i

36

2 2.5

+ 2.5

k 2.7

+ 2.3

0.5*

? 3.0

(%)

O,-Content

17.4 k 0.7

17.7 ? 0.6

16.4 k 0.6*

470 i

492 i

474 i

16.3 i- 0.5"

(vol%) PaoJF,o,

31

27

43

*P < .05: Repeated measures of ANOVA followed by Dunnett multiple as controls. To convert mm Hg to kPa, multiply the value by 0.1333.

Data are expressed

as mean

i

SD, n = 5.

498 2 41

comparisons

488 t

test. The 30-minute

42

values

are considered

194

TORBATI

Table

2. Gas Exchange (P,co,

Isocapnic-Normoxia

Variables

Time (minutes) P&O*

33.6 t 2.7

(mmHg) PdJI (mm

and 01-Carrying Capacity Before, During, = 60 mm Hg) in Anesthetized, Paralyzed,

and After Mild Hypercapnic-Normoxia and Ventilated Rats

Hypercapnic-Normoxia

0

30

ET AL

Isocapnic-Normoxia

60

90

120

33.2 2 2.8

75.8 + 8.0*

76.2 2 9.6*

32.0 2 3.8

95.8 i 8.4

94.1 -t 3.7

86.2 i 8.4

87.0 + 8.2

95.4 i 8.4

7.400 i 0.04 20.4 t 1.9

7.393 2 0.03 19.7 ir 1.4

7.135 k 0.03” 23.9 + 1.4*

7.129 i 0.04x 23.7 + 1.3”

7.391 i 0.02 18.8 + 1.7

13.9 -t 0.9

13.4 t 0.8

13.4 k 1.1

13.1 t- 1.1

12.7 5 0.9”

94.1 + 1.6

93.7 + 1.0

77.9 ? 6.1”

79.5 t 6.0*

94.0 f

17.7 ? 1.4

17.0 -t 1.1

14.2 + 2.2*

14.2 2 2.0*

16.0 2 1.5

Hg)

PH [HCOs-I, (mmol/L) Hb (g/dL) Hb ~ 02 (%) 0,.Content (vol %) Paoz/Fioz

460 2 40

*P < .05: repeated controls. To convert Data are expressed

measure

452 2 18

of ANOVA

followed

465 t 50

by Dunn&t

mm Hg to kPa, multiply the value as mean 2 SD, n = 7.

multiple

comparisons

456 f 67 test.

1.4

458 t 40

The 30-minute

values

are considered

as

by 0.1333.

animals for each repeated variable. Severe hypercapnia also produced no significant changes in the Hb concentration (Table 3). Qualitatively, the results of the gas-exchange profiles during severe hypercapnia resembled those observed during mild hypercapnia, displaying a strong Bohr effect (Fig. 1). The latter created a significantly lower Hb-02, compared with both controls and mild hypercapnia (Tables 1 to 3). Consequently, the O2 content was extremely lower than those observed during either normoxic ventilation or mild hypercapnic ventila-

tion. Severe hypercapnia significantly reduced the Pao,, compared with control rats. However, the Paoz remained within a relatively normoxic range (Table 3). Severe hypercapnia did not affect the ability of oxygen to defuse across the alveolus, as there were no changes in the Pao,/Fio, ratio (Fig. 2). Eflect of Mild and Severe Hypercapnia on MABP and HR These parameters in the control group did not change significantly throughout the experiments 600 550 T

20 20

40

60

80

100

120

140

40

60

80 [H+],

100

120

140

(nmol/L)

[H+],(nmollL) Fig 1. The Bohr effect in ventilated rats subjected to mild and severe hypercapnic-normoxia causes sharp reductions in the arterial Hb-O2 with increases in the IH’I,. Data points for mild and severe hypercapnic are derived from groups II and Ill at 30 and 60 minutes of hypercapnia (n = 14 and 10, respectively). The individual data points for normocapnia are taken from blood samples in the controls (group I) at comparable periods (n = 10).

Fig 2. Arterial oxygenation, as indicated by the ratio of P,oz/Fioz, in ventilated rats subjected to mild and severe hypercapnic-normoxia. Despite a wide range of changes in [H’l,, the P,oz/Fio, ratio remains in a relatively normal range of 400 to 550. Data points for mild and severe hypercapnic are derived from groups II and Ill at 30 and 60 minutes of hypercapnia (n = 14 and 10, respectively). The individual data points for normocapnia are derived from blood samples in the controls (group I) at comparable periods (n = 10).

PERMISSIVE

HYPERCAPNIA

Table

IN VENTILATED

RATS

195

3. Gas Exchange and O,-Carrying Capacity Before, During, and After Severe fPico2 = 90 mm Hg) in Anesthetized, Paralyzed, and Ventilated

Variables

Isocapnic-Normoxia

Time

Hypercapnic-Normoxia Rats

Hypercapnic-Normoxia

30

0

Isocapnic-Normoxia

60

90

(minutes) paw (mm

Hg)

Pa02 (mm

Hg)

PH HO-,I, (mm01 Hb

33.3 ? 3.1

32.2 k 2.1

95.8 k 12.7

98.0 ? 10.2

7.373 i 0.05 18.9 i 2.1

7.386 i 18.9 i

0.04 1.5

13.2 2 0.3

12.7 t

0.5

93.3 k 2.8 16.7 2 0.3

119.7 t

24.4,

110.6 I! 21.4*

83.8 k 11.1'

82.8 k 10.4*

101.3

i- 11.0

6.994 i 0.05" 24.6 i 1.2*

7.372

k 0.08

6.962 i 24.5 i

0.05" 1.7"

30.0 -t 8.8

16.6 i 2.1

L)

(g/dL) Hb02 1%) O2 Content (vol %I

460 + 61

PaoA% 'P < .05: Repeated

measures

13.0 -t 0.7

12.5 ? 0.4

12.5 i

94.3 2 1.1

62.0 t

66.0 t

10.5*

94.9

16.2 2 0.8

10.9 t 1.8*

11.2 t

1.8"

16.0 i- 1.4

470 + 4.9 of ANOVA

followed

as controls. To convert mm Hg to kPa, multiply Data are expressed as mean t SD, n = 5.

484 + 62

by Dunnett

the value

10.8*

multiple

478 f 58

comparisons

test.

1.0

F 1.9

486 2 53

The 30-minute

values

are considered

by 0.1333.

(Figs. 3 and 4). Similarly, mild hypercapnia produced no significant changes in the MABP and HR, compared with baselines. However, the MABP was significantly reduced 30 minutes after returning to normoxic ventilation (Fig. 3). No significant changes were observed in MABP during severe hypercapnia, but it became severely depressed after return to normoxic ventilation, compared with the control group (ANOVA followed by Tukey-Kramer multiple comparison test). Severe hypercapnia, compared with both normocapnia in the same group, and the control group, generated significant increases in the HR, which were maintained up to

30 minutes after return to normoxic air ventilation (Fig. 4). DISCUSSION

Evaluation of the mechanisms of acute lung injury and repair in humans depends on the use of appropriate animal models. Extrapolation from animal studies to human critical care conditions should consider differences in comparative pathophysiology. In this study, we investigated whether hypercapnia-induced increases in Hb concentration and O+arrying capacity, observed in ventilated dogs,7,8 can also occur in anesthetized paralyzed rats, subjected to controlled ventilation. The results in rats

450-

-P

Mild

go-- -a- Controls Hypercapnia: 60--

-Z- Mild 0 Severe

*

f

Hpercapnia

f 300-

30!

0

I

30

60

Time

90

1.20

(minutes)

250 I 0

Hypercapnia 4d

30

60

Time Fig 3. Mean arterial blood pressure (MABP) in anesthetized, paralyzed, ventilated rats during normocapnia (control group), and mild and severe hypercapnia. The values at time zero are compared with successive measurements by repeated measures of ANOVA, followed with Dunnett multiple comparisons tests.

90

3 120

(minutes)

Fig 4. Heart rate (HR) in anesthetized, paralyzed, ventilated rats during normocapnia (control group), and mild and severe hypercapnia. The values at time zero are compared with successive measurements by repeated measures of ANOVA followed by Dunnett multiple comparisons test,

196

do not support our hypothesis with respect to an increase in O+arrying capacity. The Hb concentration during the first 90 minutes of the study fell in all three groups (-7% in the controls, -6% in mild hypercapnia, and -5% in severe hypercapnia). These reductions were expected, considering our previous studies, in which Hb concentration was reduced by 10% in normal male rats after removing 10 blood samples of 0.25 mL each.14In comparison, Hb concentrations in dogs were significantly increased from a control of 15.1 2 2.3 g/dL (mean ? SD; n = 8) to 16.8 + 2.1 g/dL, after stepwise increases in the inspired PcoZ from 30 to 60 mm Hg (4 to 8 kPa). At the same time the Pace, was significantly increased from a control of 40.3 2 3.6 mm Hg to 75.2 2 9.0 mm Hg.7 A similar phenomenon also occurred in dogs during 10 hours of continuous exposure to hypercapnic-normoxia (PicoZ of 60 mm Hg; 8 kPa) with maintained ventilation, showing significant increases in Hb from 14.5 ? 2.1 g/dL to 16.7 +- 1.1 g/dL (n = 6; repeated measures of ANOVA). Concurrently, the O2content was increased from 18.7 5 3.0, vol%, to 20.8 ? 1.7, and the Pacoz was increased from 36.3 -t 1.8 mm Hg to 70.3 2 5.1 mm Hg, at the 10 hours of hypercapnia.8 Our results in rats demonstrated that arterial O2 content was reduced by approximately 6% during a 90 minute study in the controls, compared with approximately 20% in mild hypercapnia and approximately 33% in severe hypercapnia, at the same period. These reductions in the O2 content were clearly the consequence of a strong Bohr effect in rats, compared with dogs.7.8In rats, the mean HbO2 was about 78%, compared with 98% in dogs at similar Pico2 of 60 mm Hg with Pacoz of 75 and 76 mm Hg, respectively.7 Thus, rats, unlike dogs, respond to hypercapnia with a strong Bohr effect without concomitant increase in Hb concentration. However, a strong Bohr effect may also facilitate the O2 unloading at the tissue level. In fact, despite drastic reductions in the arterial O2 content in rats (Tables 2 and 3), the arterial gas exchange efficiency was maintained at a relatively normal range, even during a severe hypercapnia (Fig. 2). Since arterial O2 content, in vol%, is the product of Hb-O2 and Hb, substantial desaturation of Hb-O2 severely reduces the arterial O+arrying capacity and O2 delivery. The respiratory acidosis induced reduction in the O+arrying capacity in rats could be raised

TORBATI

ET AL

by using breathing gases with higher O2 concentrations. Another possibility could be application of mild hypothermia that can both create a left shift in the Hb-O2 dissociation curve and reduce the oxygen demands.15 Rats as Experimental Models for Critical Care Conditions Recently, we have developed an experimental critical care model to simulate ventilatory requirements in spontaneously breathing anesthetized rats, subjected to acute hypoxia, hyperoxia, hypercapnia with hyperoxia,14 and hypothermia.15 The question is, are rats suitable experimental model for pulmonary critical care conditions, involving permissive hypercapnia? Our data demonstrate that hypercapnia does not increase the O+arrying capacity in ventilated rats compared with other mammals. Whether rats can benefit from a strong Bohr effect, developed during hypercapnia, remains to be studied at a tissue level. The 02-carrying capacity could be increased either by diminishing the Bohr effect or by increasing the Fioz. During hypercapnia, HbO2 could be maximized by increasing the Fioz, which may increase the potential for pulmonary oxygen poisoning. Spontaneously breathing rats subjected to both hyperoxia and hypercapnia are capable of generating a maximum Hb-O2 saturation.14 It is possible that Hb-O2 in rats will also be maximally saturated during hypercapnic-hyperoxia with controlled ventilation. However, hyperoxia alone can increase the pulmonary shunt in both humans and rats.‘6%17 Clinically, the observed decrements in O2 delivery (lower O2 content) in rats, caused by hypercapnia, needs to be weighed against the risk of VILI, known to occur with high inflation pressures in rats subjected to conventional mechanical ventilation.‘8 We suggest that rats could be suitable for critical care studies involving permissive hypercapnia under following conditions. 1. Pulmonary oxygen transport in rats subjected to hypercapnia could be improved by bufferingI or by decreasing the 2,3-diphosphoglyceride concentrations. However, this may negate the Bohr effect, reducing O2 unloading and O2 delivery to tissues. 2. Hyperoxia in combination with permissive hypercapnia can prevent Hb-OZ desaturation and reduction in O,-carrying capacity in rats, eliminating the need for buffering. This not only maintains 02-

PERMISSIVE

HYPERCAPNIA

IN VENTILATED

RATS

197

carrying capacity but also may improve O2 unloading at the tissue levels. 3. Because changes in temperature can modulate the Hb-0, saturation curve, it is possible to reduce

Hb-O2 desaturation in rats subjected to hypercapnia by mild hypothermia. The latter can significantly reduce the ventilatory requirements, without affecting arterial oxygenation.5

REFERENCES 1. Weinberger SE, Schwartzstein RM, Weiss JM: Hypercapnia. N Engl .J Med 321:1223-1230, 1989 2. Slutsky AS: Mechanical ventilation. Chest 104:18331859, 1993 3. Hickling KG, Henderson SJ, Jackson J: Low mortality associated with low volume pressure limited ventilation with permissive hypercapnia in severe adult respiratory distress syndrome. Intensive Care Med 16:372-377, 1990 4. Hickling KG, Wright T, Lauscher K, et al: Extreme hypoventilation reduces ventilator-induced lung injury during ventilation with low positive end-expiratory pressure in saline-lavaged rabbits. Crit Care Med 26:1690-1697, 1998 5. Swenson ER, Robertson HT, Hlastala P: Effects of inspired CO2 on ventilation-perfusion matching in normoxia, hypoxia, and hyperoxia. Am J Respir Crit Care Med 149: 1563.1569,1994 6. Traystman RJ, Terry PB, Menkes HA: Carbon dioxide-a major determinant of collateral ventilation. J Appl Physiol45:6974, 1978 7. Torbati D, Mangino MJ, Garcia E, et al: Acute hypercapnia increases O2 carrying capacity of the blood in ventilated dogs. Crit Care Med 26:1863-1867, 1998 8. Ramirez J, Totapally BR, Hon E, et al: Oxygen carrying capacity during prolonged hypercapnic-normoxia in ventilated dogs. Crit Care Med (in press) 9. Wetterberg T, Sjoberg T, Steen S: Effects of buffering in hypercapnia and hypercapnic hypoxemia. Acta Anaesthesiol Stand 37:343-349, 1993 10. Alexander JK, West JR, Wood JA, et al: Analysis of the

respiratory response to carbon dioxide inhalation in varying clinical states of hypercapnia, anoxia, and acidosis. J Clin Invest 34:511-532, 1954 11. Pepelko WE, Dixon GA: Arterial blood gases in conscious rats exposed to hypoxia, hypercapnia, or both. J Appl Physiol 38:581-587, 1975 12. Brackett NC, Cohen JJ, Schwartz WB: Carbon dioxide curve of normal man: Effect of increasing degree of acute hypercapnia on acid base balance. N Engl J Med 272:6-l 1, 1965 13. Cohen JJ, Brackett NC, Schwartz WB: The nature of the carbon dioxide titration curve in the normal dog. J Clin Invest 43~777.786, 1964 14. Torbati D, Ramirez J, Hon E, et al: Experimental critical care in rats: Gender differences in anesthesia, ventilation and gas-exchange. Crit Care Med 27:1878-1884, 1999 15. Torbati D, Camacho MT, Hultquist K; et al: Ventilation and gas-exchange responses to hypothermia and pentobarbital anesthesia in female rats. Crit Care Med 26:A-119, 1998 16. Wagner PD, Laravuso R, Uhl R, et al: Continuous distributions of ventilation-perfusion ratios in normal subjects breathing air and 100% oxygen. J Clin Invest 53:54-68, 1974 17. Truog WE, Hlastala MP, Standaert TA, et al: Oxygeninduced alteration of ventilation-perfusion relationship in rats. J Appl Physiol 47:1112-1117, 1979 18. Dreyfuss D, Soler P, Basset G, et al: High inflation pressure pulmonary edema. Am Rev Respir Dis 137:1159-1164, 1988