History

A 12-year-old spayed female Staffordshire Terrier mixed breed weighing 33.6 kg (74.0 lb) was evaluated for removal of 2 subcutaneous masses over the left flank and medial aspect of the left tarsus. Previously, cytologic examination of fine-needle aspirates had tentatively identified the masses as spindle cell tumors. The patient had a long history (> 2 years) of respiratory tract disease characterized clinically by intermittent coughing with occasional exercise intolerance and tachypnea. Serial radiography revealed a progressive, moderate-to-severe, diffuse bronchointerstitial pattern consistent with chronic allergic or idiopathic bronchitis or possibly fibrosis secondary to previous infectious disease (Figure 1). Clinical improvement was noted with administration of fluticasone propionate (13.1 μg/kg [5.9 μg/lb], by inhalation, q 12 h) and theophylline (8.9 mg/kg [3.9 mg/lb], PO, q 12 h).

View Full Size

Right lateral (A) and ventrodorsal (B) radiographic views of the thorax obtained prior to anesthesia in a 12-year-old Staffordshire Terrier mixed breed with chronic, progressive respiratory disease. A diffuse moderate-to-severe bronchointerstitial pattern is present.

Citation: Journal of the American Veterinary Medical Association 239, 4; 10.2460/javma.239.4.442

• Download Figure
• Download figure as PowerPoint slide

Echocardiography had been performed to rule out pulmonary hypertension and had revealed aortic root dilatation, severe left ventricular hypertrophy, and a small amount of tricuspid regurgitation. Pulmonary hypertension was judged unlikely on the basis of pulmonary arterial pressures estimated by use of spectral Doppler ultrasonographic interrogation of peak tricuspid regurgitation velocity as well as the lack of right ventricular hypertrophy or pulmonary arterial enlargement. Mild systemic hypertension (systolic arterial pressure, 160 to 180 mm Hg) was confirmed by means of repeated Doppler ultrasonographic monitoring, and amlodipine (0.15 mg/kg [0.07 mg/lb], PO, q 12 h) and atenolol (0.19 mg/kg [0.08 mg/lb], PO, q 12 h) were prescribed. The dilatation of the aortic root was consistent with systemic hypertension, but the degree of left ventricular hypertrophy was judged to be disproportionate in nature. The findings were consistent during repeated echocardiography and were considered unrelated to the pulmonary disease.

On physical examination, the patient was bright, alert, and responsive. A mild sinus arrhythmia with strong and synchronous femoral pulses was noted. The patient was panting, and increased bronchovesicular sounds were auscultated over all lung fields. All values on a recent CBC and serum biochemistry panel, including measurement of electrolyte concentrations, were within reference ranges. While the patient was conscious and breathing room air, an arterial blood sample was obtained from the dorsal pedal artery and submitted for blood gas analysis (Table 1). A patient-side handheld analyzer^a was used, and all results were corrected to the patient's body temperature.

The patient was premedicated with hydromorphone (0.1 mg/kg [0.045 mg/lb], IV), and oxygen was provided by mask for approximately 5 minutes prior to induction of anesthesia. A Doppler ultrasound flow probe was placed over the palmar metatarsal artery, and a lead II ECG was monitored. Anesthesia was induced with ketamine (1.1 mg/kg [0.5 mg/lb], IV), followed by propofol (2.3 mg/kg [1.0 mg/lb], IV, to effect). A 12-mm internal diameter silicone cuffed orotracheal tube was placed, and an adult circle breathing circuit with a 3-L rebreathing bag was attached. Anesthesia was maintained with sevoflurane in oxygen; the initial vaporizer setting was 4% with an oxygen flow rate of 1.5 L/min. The patient was breathing spontaneously, but manual intermittent positive pressure ventilation with a peak inspiratory pressure of 15 cm H₂O was initiated to maintain Petco₂ between 35 and 45 mm Hg.

A 20-gauge, 1.25-inch catheter was placed in the right dorsopedal artery to allow measurement of arterial blood pressure and collection of samples for arterial blood gas analysis. Arterial line tubing was connected from the catheter to a disposable transducer placed at the level of the right atrium and zeroed to atmospheric pressure. The transducer was attached to a multiparameter anesthesia monitor.^b In addition, results of pulse oximetry and capnography were monitored continuously. Approximately 10 minutes after anesthesia was induced, the respiratory rate was 12 breaths/min; heart rate was 114 beats/min; Petco₂ was 45 mm Hg; systolic, diastolic, and mean arterial blood pressures (measured directly) were 105, 75, and 85 mm Hg, respectively; and body temperature was 36.5°C (98°F). The Spo₂ was 93%. An arterial blood sample was collected at this time and submitted for blood gas analysis (Table 1).

Question

What is the interpretation of the results of blood gas analysis of the arterial blood sample collected 10 minutes after induction of anesthesia in this patient? What treatment, if any, is indicated?

Answer

Mild metabolic (or nonrespiratory) acidosis and relative hypoxemia are present. Respiratory compensation for the metabolic acidosis has not occurred, likely because of the confounding effects of anesthetic drug administration and manual positive-pressure ventilation. Although not immediately life-threatening, the relative hypoxemia is severe, indicates a substantial disturbance in gas exchange, and should be addressed promptly.

Intermittent mandatory ventilation was initiated with a time-cycled, pressure-limited anesthesia ventilator^c with a tidal volume of 10 mL/kg (4.5 mL/lb) and respiratory rate of 15 breaths/min. Peak inspiratory pressure was 15 cm H₂O as indicated by a manometer. In addition, a unidirectional 7.5 cm H₂O PEEP valve was placed in the expiratory limb of the breathing circuit. After 10 minutes, a third arterial blood sample was obtained for blood gas analysis, which revealed an improvement in blood oxygenation and development of mild respiratory acidosis that was considered tolerable (Table 1).

Discussion

Arterial hypoxemia denotes low Pao₂. This differs from hypoxia, which refers to low partial pressure of oxygen at the tissue level and is currently difficult to measure clinically. Absolute hypoxemia is generally defined as a Pao₂ ≤ 60 mm Hg. Because of the shape of the oxygen-hemoglobin dissociation curve, this is equivalent to an Spo₂ of approximately 90%. Importantly, however, pulse oximeter algorithms are based on the relationship between hemoglobin and oxygen in humans, not dogs. The conventional method for discussing the affinity of hemoglobin for oxygen is the partial pressure of oxygen at which 50% of hemoglobin is bound. This value is slightly higher for canine hemoglobin than for human hemoglobin. Thus, in dogs, measured Spo₂ is expected to be somewhat low for any given Pao₂,¹ as exemplified by results for the dog described in the present report. In this dog, the measured Spo₂ was 93% when the Pao₂ was 89 mm Hg, whereas the expected Spo₂ reading for this Pao₂ would be approximately 96%. Other factors can also affect the position of the oxygen-hemoglobin dissociation curve, altering measured Spo₂ relative to Pao₂, including temperature, acid-base balance, and 2,3-diphosphoglycerate concentration.

The dog described in the present report was classified as being relatively hypoxemic following induction of anesthesia. That is, the Pao₂ was unexpectedly low, given the Pio₂. This was also the case prior to induction of anesthesia when the patient was breathing room air. Pulmonary gas exchange is evaluated via calculation of the alveolar-arterial difference in partial pressure of oxygen (ie, the difference between Pao₂ and Pao₂), where Pao₂ = Pio₂ − (Paco₂/respiratory quotient). The Pio₂ is dependent on Fio₂, atmospheric pressure, and the partial pressure of water vapor (ie, Pio₂ = Fio₂ × [atmospheric pressure − partial pressure of water vapor]). In a dog with normal lung function that is breathing room air (ie, Fio₂ = 21%; Pio₂ at sea level =150 mm Hg), the alveolar-arterial difference should be < 10 mm Hg. The alveolar-arterial difference is much larger if the Fio₂ is > 21%, and several equations have been proposed for evaluating lung function under such conditions.² A simple and commonly used option is the Pao₂:Fio₂ ratio. Although this calculation ignores the contribution of hypoventilation (in that Paco₂ is not part of the equation), with a high Fio₂, hypoventilation would have to be severe to substantially affect Pao₂. A normal Pao₂:Fio₂ ratio is approximately 500, whereas in the present case, the calculated Pao₂:Fio₂ ratio 10 minutes after induction of anesthesia was approximately 94, indicating a severe gas exchange problem.

Several assumptions were made when calculating the alveolar-arterial difference in the dog described in the present report. Atmospheric pressure changes with elevation and weather conditions and was not measured but was, rather, assumed to be 760 mm Hg on the basis of the hospital's location. In addition, Paco₂ was used to calculate Pao₂, rather than Paco₂. Because CO₂ is highly diffusible and measurement of Paco₂ is difficult in clinical situations, Paco₂ is generally presumed to reflect Paco₂, although this may not be true in dogs with severe pulmonary disease.² The value used for the respiratory quotient could also have affected the calculated alveolar-arterial difference. The respiratory quotient is the ratio of O₂ consumption to CO₂ production and can vary with diet, disease, and Pio₂. For the dog described in the present report, the respiratory quotient was assumed to be 0.8 when the patient was breathing room air.² Finally, the partial pressure of water vapor varies with body temperature but the change is rarely substantial enough to matter when calculating Pao₂.³

There are 5 potential causes of hypoxemia: low Pio₂, hypoventilation, ventilation-perfusion mismatch in the lungs, anatomic (ie, cardiac) or physiologic (ie, pulmonary) right-to-left shunting, and diffusion impairment. When relative or absolute hypoxemia occurs, potential causes should be investigated and addressed as necessary.

Values of Fio₂ were not measured in the dog described in the present report, and it is possible that a dysfunction of O₂ supply could have contributed to the hypoxemia that was detected during anesthesia. If hypoxemia is documented, the fresh gas supply and breathing circuit should be inspected quickly to ensure appropriate flow. At the time this dog was anesthetized, the hospital oxygen supply consisted of a manifold of size H gas cylinders, which should deliver 100% medical-grade oxygen. The oxygen flow meter on the anesthetic machine indicated appropriate gas flow, the breathing circuit was connected correctly to the anesthetic machine and endotracheal tube, and the hospital oxygen supply line pressure was confirmed to be 48 psi, making low Pio₂ an unlikely cause of the dog's hypoxemia. It is possible that Pio₂ could have been low because of a high time constant for the anesthetic breathing circuit. The time constant is defined as capacity of the anesthetic breathing circuit (L) divided by the fresh gas flow (L/min), and 3 time constants are required to achieve 95% of desired fresh gas concentration. Assuming a capacity of 5 L in the circle breathing circuit that was used and a fresh gas flow of 1.5 L/min, the oxygen concentration would have reached approximately 95% at 10 minutes. Therefore, low Pio₂ was unlikely to have been the cause of the relative hypoxemia documented at the time of the blood gas analysis collected 10 minutes after anesthetic induction.

At an Fio₂ close to 100%, hypoventilation would have to be severe to be the sole cause of hypoxemia. Hypoventilation did not appear to contribute to the relative hypoxemia seen in the dog described in the present report, because the patient was ventilating spontaneously with occasional manual positive pressure ventilation, and Petco₂ and Paco₂ were within reference ranges. Apnea is an extreme form of hypoventilation and can cause hypoxemia; apnea should be addressed by manually or mechanically ventilating the patient.

The most common causes of hypoxemia during anesthesia are ventilation-perfusion mismatch and physiologic right-to-left shunting secondary to atelectasis. Increased scatter of ventilation-perfusion relationships indisputably occurs during anesthesia⁴; however, an Fio₂ of close to 100%, as is typically used in veterinary anesthesia, is adequate to compensate for areas of low ventilation-perfusion. Intrapulmonary shunting, such as that caused by atelectasis, can be considered an extreme form of ventilation-perfusion mismatch, and increasing Fio₂ cannot offset the resulting decreases in Pao₂. Atelectasis occurs quickly during anesthesia as a result of physical compression of the lungs and preferential absorption of oxygen as nitrogen is washed out of the alveoli.⁵ The patient described in the present report had preexisting chronic pulmonary disease characterized by defective gas exchange, as evidenced by results of blood gas analysis performed while the dog was breathing room air, and it is likely that anesthesia with a high Fio₂ aggravated the pulmonary abnormalities. Ventilation-perfusion mismatch can also be caused by endobronchial intubation. In such a circumstance, one lung is not ventilated at all but perfusion of that lung continues. Endobronchial intubation can be prevented by ensuring that the tip of the endotracheal tube is palpable at the thoracic inlet, which was the case in this patient. Anatomic (intracardiac) right-to-left shunting was ruled out in this case by repeated echocardiography prior to anesthesia.

Diffusion impairment is uncommonly the sole cause of hypoxemia because it requires thickening or destruction of the alveolar-capillary membrane, as occurs with diffuse interstitial fibrosis or emphysema.⁶ Some authors include pulmonary edema as an etiology of a thickened alveolar-capillary membrane and resultant diffusion impairment. However, diffusion impairment is rarely the primary limiting factor in patients with hypoxemia and is usually corrected by use of high Fio₂. The exact nature of the pulmonary disease in this patient had not been characterized, and it is possible that diffusion impairment could have contributed to the hypoxemia.

To decrease ventilation-perfusion mismatch and physiologic shunting in the dog described in the present report, mechanical ventilation was instituted and a 7.5 cm H₂O PEEP valve was inserted into the breathing circuit. In addition, the ventilator used in this dog had a standing bellows, which applies 2 to 3 cm H₂O of PEEP, owing to the weight of the bellows. Application of PEEP maintains positive airway pressure relative to ambient pressure during expiration, thereby stabilizing alveoli that would otherwise collapse and increasing functional residual capacity. This generally results in improved gas exchange, as evidenced in this case by improvements in Pao₂ and the Pao₂:Fio₂ ratio by the time the third blood sample was collected for blood gas analysis. However, PEEP can have deleterious effects on cardiac output by decreasing venous return to the heart.⁷ Decreased cardiac output in the face of a large physiologic shunt will further decrease Pao₂. In fact, PEEP alone has not been demonstrated to consistently improve Pao₂ in humans anesthetized with an Fio₂ of approximately 50%, possibly because of regional redistribution of pulmonary blood flow and increased ventilation-perfusion scatter.⁸ In addition, application of PEEP can increase dead space ventilation by splinting open alveoli that are not well perfused. The optimal level of PEEP is that which improves oxygenation without causing further pulmonary or hemodynamic compromise, and PEEP should be applied cautiously in patients with preexisting cardiovascular instability, such as hypovolemia or profound vasodilation. In the patient of this report, PEEP did not negatively affect cardiovascular parameters, as evidenced by maintenance of adequate blood pressures.

Recruitment maneuvers can be performed to re-expand collapsed alveoli in healthy and diseased lungs. This technique consists of transient increases in peak inspiratory pressure and sometimes PEEP in an attempt to reopen (or recruit) collapsed alveoli. A variety of recruitment maneuvers have been described, and the optimum strategy has not yet been defined. A common practice is that of inflation whereby the peak inspiratory pressure is increased in a stepwise fashion to 40 to 50 cm H₂O and held for 30 to 60 seconds at a time.⁹ Continued use of PEEP must follow a recruitment maneuver to avoid recurrence of atelectasis. Owing to the unknown nature of the preexisting pulmonary disease in the patient of this report, it was decided to perform a recruitment maneuver only if application of intermittent mandatory ventilation and PEEP did not adequately improve oxygenation parameters.

The protocol for anesthetic induction in the dog described in the present report consisted of IV administration of a low dose of ketamine followed by IV administration of propofol until the patient could be intubated. This induction protocol was chosen because ketamine increases sympathetic outflow, and its use prior to propofol may prevent the hypotension that can follow administration of an induction dose of propofol.¹⁰ Ketamine is also a superb bronchodilator, has analgesic and antihyperalgesic effects, and reduces perioperative opioid requirements.^11,12 In the dog described in the present report, PEEP was decreased to 5 cm H₂O near the end of surgery; mechanical ventilation was then discontinued, and a nasal O₂ catheter was placed. The patient recovered without complications and without clinical signs of respiratory complications. The patient was alive 60 days after surgery, and clinical signs related to the chronic respiratory disease were static.