History

A 4-year-old 11.2-kg (24.7-lb) castrated male French Bulldog was evaluated on an emergency basis at the University of Minnesota Veterinary Medical Center because of acute bilateral hind limb paraparesis. The patient became nonambulatory in the hind limbs after having been dropped by the owner. Past history included inflammatory bowel disease controlled with dietary modifications and a greenstick tibial fracture of unspecified etiology as a puppy. More recently, the patient had been acting lethargic with episodes of trembling for several days prior to the onset of paresis. It also had had a decreased appetite and 1 episode of diarrhea.

Initial physical examination revealed the patient to be anxious, alert, and responsive. The dog was slightly obese with a body condition score of 7 on a scale from 1 to 9. Rectal temperature was 40.4°C (104.9°F). Tachycardia (HR, 200 beats/min) and synchronous femoral pulses were noted. Results of thoracic auscultation were unremarkable; however, the patient was panting. Abdominal palpation yielded no clinically important findings. The dog was nonambulatory with hind limb paraparesis; the left hind limb was weaker than the right. Signs of pain were evident during palpation of the caudal portion of the thoracic spine. The urinary bladder was small on palpation, and urine could be expressed. Results of the remainder of the physical examination were unremarkable.

Biochemical testing was performed. Abnormalities included high total plasma protein concentration (7.0 g/dL; reference range, 5 to 6.9 g/dL), high albumin concentration (4.0 g/dL; reference range, 2.7 to 3.7 g/dL), and low serum phosphorus concentration (1.5 mg/dL; reference range, 3.3 to 6.8 mg/dL). Results of a CBC were unremarkable except for slight hemolysis with clumped platelets.

Survey radiography of the thoracolumbar vertebral column revealed evidence suggestive of IVDD, and advanced imaging was recommended. Given that the patient had signs of central recognition of deep pain as well as motor function on initial examination, imaging was scheduled for the following day.

Hydromorphone (0.1 mg/kg [0.045 mg/lb], IM) and prednisone (0.22 mg/kg [0.1 mg/lb], PO) were administered shortly after admission to the hospital.

The initial hyperthermia and tachycardia, attributed to stress and pain, had resolved by the time of the preanesthetic examination the next morning. However, neurologic status had worsened overnight, and no signs of deep pain sensation were elicited. Rectal temperature was 38.9°C (102°F), HR was 120 beats/min, and the patient was panting. The patient was premedicated with acepromazine (0.013 mg/kg [0.006 mg/lb], IM) and hydromorphone (0.18 mg/kg [0.08 mg/lb], IM). A 22-gauge right cephalic catheter was placed IV, and anesthesia was induced with diazepam (0.18 mg/kg, IV) and propofol (3.6 mg/kg [1.6 mg/lb], IV, to effect). A 6-mm cuffed endotracheal tube was placed, and anesthesia was maintained with sevoflurane in oxygen (500 mL/min), delivered through a semiclosed circle system. Lactated Ringer's solution (10 mL/kg/h [4.5 mL/lb/h], IV) was administered throughout the anesthetic episode.

Following induction of anesthesia, the patient was transported to the diagnostic imaging suite and positioned in dorsal recumbency for MRI. Anesthetic monitoring, with an MRI-compatible monitor,^a included continuous assessment of HR, RR, Spo₂, Petco₂, and an ECG, and indirect oscillometric measurement of blood pressure with the cuff placed proximal to the left carpus. Because of the inaccessibility of the patient during the imaging procedure, assessment of physical variables was limited. The patient was mechanically ventilated^b with a tidal volume of 200 mL (17.8 mL/kg [8 mL/lb]) to a peak airway pressure of 10 cm HO at a rate of 12 breaths/min. End-tidal partial pressure of carbon dioxide, measured during imaging, was maintained between 25 and 29 mm Hg.

Approximately 5 minutes after the start of the MRI procedure (20 minutes after anesthetic induction), the patient became mildly bradycardic (HR, 59 beats/min) and mildly hypotensive (SAP, 85 mm Hg; MAP, 59 mm Hg; DAP, 39 mm Hg). A dose of glycopyrrolate (0.004 mg/kg [0.002 mg/lb], IV) was administered, which improved the bradycardia (HR, 72 beats/min) and resolved the hypotension (SAP, 102 mm Hg; MAP, 70 mm Hg; DAP, 45 mm Hg). Approximately 25 minutes later, the HR decreased to 69 beats/min and blood pressure also decreased (SAP, 79 mm Hg; MAP, 67 mm Hg; DAP, 46 mm Hg). A second dose of glycopyrrolate (0.004 mg/kg, IV) was administered, which increased the HR to 85 beats/min. Blood pressure also increased (SAP, 99 mm Hg; MAP, 75 mm Hg; DAP, 52 mm Hg).

Magnetic resonance imaging revealed left-sided dorsal intervertebral disk herniation at T13-L1 and L1–2. Spinal cord changes at these sites were indicative of possible severe edema or early myelomalacia. Multiple other sites of IVDD without spinal cord compression were also seen.

Approximately 90 minutes after induction of anesthesia, the patient was moved to the surgical suite and positioned in sternal recumbency for left hemilaminectomy at T13-L1 and L1–2. Anesthetic monitoring^c during surgery included continuous assessment of HR, RR, Spo₂,^d and an ECG, and indirect oscillometric measurement of blood pressure with a No. 3 cuff placed proximal to the left carpus. The patient was administered a preoperative dose of cefazolin (22.3 mg/kg [10.2 mg/lb], IV). Additional doses of cefazolin were administered IV every 90 minutes throughout the procedure, for a total of 3 doses. A CRI of fentanyl (13.4 μg/kg/h [6 μg/lb/h], IV) was administered via a syringe pump^e for perioperative analgesia.

Body temperature at the start of surgery was 34.5°C (94.1°F) and improved to 38.7°C (101.6°F) with supplemental heat from a forced warm air blanket^f and circulating warm water blanket.^g Anesthesia was maintained with a delivered concentration of sevoflurane of 2.75% to 4% throughout the surgical procedure, and the patient was allowed to breathe spontaneously. The RR was 12 to 23 breaths/min. Oxygen saturation determined by means of pulse oximetry was 100% throughout the entire anesthetic episode. Following the initial bradycardia, which resolved with glycopyrrolate administration, the HR was consistently 85 to 100 beats/min with a normal sinus rhythm. Mucous membrane color remained pink, and capillary refill time was 1 second. Systolic arterial blood pressure, MAP, and DAP were 75 to 130 mm Hg, 64 to 91 mm Hg, and 30 to 65 mm Hg, respectively. Toward the end of the procedure, the patient became mildly hypotensive with the lowest SAP of 75 mm Hg, MAP of 55 mm Hg, and DAP of 30 mm Hg. A reduction in delivered sevoflurane concentration as well as administration of a bolus of hetastarch (1.8 mL/kg [0.81 mL/lb], IV) only mildly improved the blood pressure. Because the surgeon was in the process of closing the surgical incision, a clinical judgment was made to not pursue more aggressive forms of blood pressure management. The CRI of fentanyl was discontinued at the end of the surgical procedure. Total imaging time was 1 hour, total surgical time was 2.6 hours, and total anesthesia time was 4.75 hours. The additional anesthesia time was attributed to transport to and from the MRI suite and surgical preparation of the patient.

During anesthetic recovery, the patient became tachypneic while still intubated and subjectively appeared distressed. Postoperative body temperature was 38.7°C (101.6°F), and HR at the conclusion of the surgical procedure was 120 beats/min. The patient was assessed as painful, and additional analgesia was thought necessary to make the patient more comfortable. A dose of hydromorphone (0.09 mg/kg [0.04 mg/lb], IV), half of what had been given IM as a premedicant, was administered as the patient was extubated. Immediately, the patient stopped breathing, the mucous membranes and tongue became cyanotic, no pulses were palpable, and the palpebral reflex was lost. The patient was immediately reintubated and connected to a semiclosed circle system; 100% oxygen was delivered with intermittent positive pressure ventilation at a peak pressure up to 20 cm H₂O and a rate of 6 to 10 breaths/min. Administration of lactated Ringer's solution (10 mL/kg/h, IV) was continued. Atropine (0.02 mg/kg [0.009 mg/lb], IV) and naloxone (0.02 mg/kg, IV) were administered, and closed chest compressions at a rate of 80 to 100 compressions/min were performed with the patient in left lateral recumbency. Open chest cardiac massage was not pursued owing to the size of the patient. Electrocardiogram^h leads were placed, heart rhythm was monitored, and a Dopplerⁱ probe was placed over the right eye. No response was evident on the ECG; therefore, epinephrine (0.09 mg/kg, IV) and vasopressin (0.09 U/kg, IV) were administered. Ventricular fibrillation was noted on the ECG, and external defibrillation^h (4.5 J/kg [2 J/lb]) was performed. Chest compressions were resumed, but ventricular fibrillation was still noted on the ECG. A second dose of epinephrine (0.09 mg/kg, IV) was administered, and the patient was again externally defibrillated (4.5 J/kg). Chest compressions were resumed, and a normal sinus rhythm was noted on the ECG. Time from initial arrest to successful resuscitation was 10 minutes.

Approximately 7 minutes after initial resuscitation and with no provocation, while the patient was lying quietly, profound bradycardia followed by cardiac arrest developed. Atropine (0.02 mg/kg, IV), epinephrine (0.09 mg/kg, IV), and sodium bicarbonate (1.07 mEq/kg [0.49 mEq/lb], IV) were administered, and chest compressions were resumed. Ventricular fibrillation was evident on the ECG, and the patient was externally defibrillated (4.5 J/kg). Chest compressions were resumed, but ventricular fibrillation was still evident. The patient was again externally defibrillated (4.5 J/kg), followed by chest compressions. Sinus tachycardia was evident on the ECG, and the patient began to breathe spontaneously. Time from the second arrest to successful resuscitation was 8 minutes. Total time the patient was clinically dead was approximately 18 minutes.

Because the patient had profound bradycardia and had had 2 episodes of cardiac arrest, a CRI of epinephrine (0.1 μg/kg/min, IV) was begun to augment HR and contractility in the short term,¹ but was discontinued after the tachycardia became worse. A CRI of dobutamine (4.4 μg/kg/min [2.0 μg/lb/min], IV) was instituted for inotropic support,² and a CRI of dopamine (2.2 μg/kg/min [1.0 g/lb/min], IV) was administered to potentially improve cardiac contractility and stimulate dopaminergic receptor activity in the kidney,³ even though strong evidence supporting this decision does not exist. Mannitol (0.5 g/kg [0.23 g/lb], IV) was administered slowly over 20 minutes to counteract potential postar-rest cerebral edema and reperfusion injury.³ A venous blood gas sample obtained from the lingual vein soon after recovery from the second arrest was analyzed,^j with adjustments made for the patient's temperature of 35°C (95°F). Abnormalities included a pH of 7.109 (reference range, 7.30 to 7.47), Pvco₂ of 60.9 mm Hg (reference range, 28.9 to 44.4 mm Hg), base excess of −10 mmol/L (reference range, −7.4 to 2.8 mmol/L), and lactate concentration of 8.79 mmol/L (reference range, 0.21 to 3.80 mmol/L). Other results were within reference limits, including Pvo₂ of 44.0 mmol/L (reference range, 27.4 to 56.0 mmol/L) and bicarbonate concentration of 19.8 mmol/L (reference range, 17.8 to 27 mmol/L).

Given the venous blood gas analysis results indicative of hypercapnia and the previous administration of sodium bicarbonate, the patient was connected to a mechanical ventilator^k for treatment of the respiratory acidosis. Heart rate was 160 beats/min, RR was 43 breaths/min (spontaneous and controlled breaths), and Petco₂¹ was 25 mm Hg. A second venous blood sample was obtained in a similar manner as the first, approximately 45 minutes after the second arrest. The pH had improved to 7.251, and Pvco₂ had improved to 33.3 mm Hg; however, bicarbonate concentration had decreased to 14.7 mmol/L, base excess had decreased to −13 mmol/L, and lactate concentration had increased to 10.74 mmol/L. The clinical assessment at this time was that improvements in perfusion had allowed by-products of anaerobic glycolysis (ie, lactate) to be carried into the circulation,³ explaining the increase in lactate concentration and decrease in bicarbonate concentration. It was also considered possible that vasoconstriction and perfusion status had worsened and that administration of additional fluids was indicated. On the basis of the assessment of the patient and the overall clinical improvement, it was decided to continue with the treatment plan and continue IV fluid treatment with lactated Ringer's solution (10 mL/kg/h, IV). Multiple attempts were made to place an arterial catheter in the dorsal pedal arteries but were unsuccessful. A surgical cut down over the left femoral artery was performed, and an indwelling arterial catheter was placed to allow direct measurement of blood pressure and to obtain serial arterial blood gas samples once the patient was moved to the intensive care unit.

The patient was disconnected from the ventilator and allowed to breathe 100% oxygen spontaneously because it continued to have tachypnea and asynchrony with the ventilator and the Pvco₂ had improved. An arterial blood sample was obtained approximately 60 minutes after the second arrest and analyzed (pH, 7.244; Paco₂, 42 mm Hg; Pao₂, 568 mm Hg; ionized calcium concentration, 4.4 mg/dL [reference range, 5.1 to 5.9 mg/dL]); lactate concentration was not measured. Given the low ionized calcium concentration, calcium gluconate (0.4 mEq/kg [0.18 mEq/lb], IV) was administered over 20 minutes. At that time, HR was 162 beats/min, RR was 90 breaths/min, and Petco₂ was 20 mm Hg.

The patient regained a palpebral reflex shortly after the second successful resuscitation. A swallow reflex was present approximately 90 minutes after arrest, and extubation was performed without further incident. The patient was transferred to the intensive care unit for continued care and monitoring and received additional doses of hydromorphone (0.05 mg/kg [0.023 mg/lb], IV) every 6 hours for the first 24 hours after surgery for pain management. The decision to continue use of a full μ-opioid receptor agonist as an analgesic was made on the premise that the patient had just had a major surgical procedure as well as chest compressions on 2 separate occasions. It was also thought that the patient would be unlikely to have another vagally mediated event secondary to hydromorphone administration alone because it had been administered this particular opioid on several occasions during this hospitalization. Serial blood gas analyses performed every 4 hours revealed that lactate concentration was within reference limits 8 hours after arrest.

The patient was able to see and had normal mentation the following day. Postoperative pain management was changed from hydromorphone to buprenorphine (0.01 mg/kg [0.0045 mg/lb], IV, q 8 h). A corneal ulcer was evident in the right eye, and triple antimicrobial ointment was administered in the affected eye every 8 hours. Neurologic assessment of the hind limbs at discharge revealed conscious proprioception deficits, but motor function was present in both hind limbs. The patient was bright, alert, and responsive and was discharged to the owners for continued care at home 3 days after arrest. The patient continued to do well, and at its 2-month reevaluation had slight conscious proprioception deficits in the hind limbs, some hind limb weakness when walking, and slight bilateral muscle atrophy in the hind limbs, which were attributed to the IVDD.

Question

What was the cause of this patient's arrest at the time of anesthetic recovery?

Answer

Brachycephalic dogs have a tendency to have a higher vasovagal tone than do dogs of other breeds. Thus, in this dog, concurrent opioid administration and tracheal extubation may have resulted in a substantial vagally mediated response leading to cardiopulmonary arrest.

Discussion

Vagally mediated arrest or asystole is a rare phenomenon, with little published information on this condition in humans and animals. Vagally mediated bradycardia⁴ is more common. Parasympathetic or vagal innervation of the heart is primarily to the sinoatrial and atrioventricular nodal tissue.⁵ At the receptor level, acetylcholine acts on M₂ muscarinic receptors that activate 1_K/Ach (an acetylcholine-activated subtype of inward rectifying current) on sinoatrial and atrioventricular nodal tissue cells. This leads to membrane hyperpolarization, which decreases the rate of spontaneous depolarization, causing a decrease in HR.⁶ Development of bradycardic complications leading to full arrest or asystole seems to be a common clinical situation encountered among anesthesiologists; however, published references in the veterinary literature are lacking, and the issue of vagally mediated arrest as an adverse effect of opioid administration or endotracheal intubation is only briefly mentioned in anesthesiology textbooks.^7–9

Opioid receptor agonists, especially when administered IV as a bolus or CRI, can cause relative bradycardia in any patient.^8,9 This bradycardia can be transient or can be prolonged and sometimes quite severe. Rats receiving morphine developed bradycardia for several minutes following an IV bolus injection, but bilateral cervical vagotomy blocked the bradycardia as completely as did naloxone or atropine.¹⁰ Similarly, opioid-induced bradycardia was found to be vagally mediated in dogs given fentanyl IV.¹¹ These studies validate the idea that opioid-induced bradycardia is due to stimulation of vagal afferents.

The dog described in the present report was administered an IV bolus of hydromorphone, a full μ-opioid receptor agonist,⁸ at the time of anesthetic recovery, which could be expected to have caused at least transient bradycardia. Even though the dog was given glycopyrrolate early in the anesthetic episode, this anti-cholinergic is eliminated between 30 minutes and 3 hours after IV administration.¹² Therefore, at the time of arrest (4 hours after glycopyrrolate administration), the glycopyrrolate would likely have been eliminated and would not have attenuated any vagal response induced by the hydromorphone.

Possibly exacerbating the bradycardic response to hydromorphone administration was the fact that this patient likely had a therapeutic plasma concentration of fentanyl¹³ in the immediate postoperative period. The CRI of fentanyl administered throughout the surgical period had been discontinued for only approximately 15 minutes before the first cardiac arrest occurred. However, the patient was considered to be painful at the time of recovery owing to a combination of tachypnea and distress. This could potentially have been related to borderline hyperthermia because the dog's postoperative body temperature was 38.7°C (101.6°F) and opioids are known to alter the hypothalamic thermoregulatory system, although the mechanism by which they do so remains unknown.⁸ Differentiating pain from dysphoria can be difficult in clinical cases, and the potential exists that this patient had dysphoria rather than pain at the time of anesthetic recovery. In this case, the additional dose of hydromorphone further decreased sympathetic tone, worsened vagally mediated bradycardia, and led to asystole. The concurrent timing of opioid administration with extubation may or may not have been important in this case. Although this patient received opioids at the time of extubation, we frequently administer opioids at induction to facilitate intubation. The argument could be made that the hydromorphone administered IV was not a factor in the development of cardiac arrest because the patient was extubated at the same time as the drug was administered.

In human patients, asystole has been documented following the administration of remifentanil, sufentanil, or fentanyl at anesthetic induction.^14,15 This complication has been attributed to a decrease in sympathetic tone and concurrent vagally mediated bradycardia. In other case reports,^16–19 patients received opioids at induction but had extreme bradycardia that progressed to asystole only when laryngoscopy was attempted. In case reports describing human patients with arrest, opioids were given several minutes prior to development of asystole, allowing time for the drug to reach appropriate receptor sites and induce an effect.

Extubation was more likely responsible for the sudden cardiac arrest in the patient described in the present report. Tracheal manipulation during intubation or extubation typically generates a sympathetic response but can also cause a parasympathetic response.²⁰ Depending on the individual patient's reaction, this can lead to a brief alteration in HR. More severe responses include tachycardia, bradycardia, cardiac dysrhythmias, and cardiac arrest.²⁰ For this particular patient, given the recent opioid administration IV followed by immediate tracheal extubation, it is possible there was a compounded parasympathetic effect.

Brachycephalic dogs have been shown to have a higher resting vasovagal tonus index than dogs of other breeds, but the mechanism of this is unknown at this time.²¹ The vasovagal tonus index is calculated from the natural logarithmic fluctuation in R-R interval over 20 heart cycles on an ECG. This number is then used to evaluate HR variability in dogs.²¹ The higher the vasovagal tonus index, the more HR variability the patient has, meaning the higher parasympathetic effects on the heart. Given this patient's breed (French Bulldog), it is possible that it had a high vasovagal tonus index and thus may have been more susceptible to the vagal effects of opioid administration as well as the vagal effects of tracheal stimulation during extubation.²²

Overall anesthetic-related death rates in dogs and cats have been estimated to range from 0.1% to 2%, depending on the health of the animal and the institution.²³ These rates are still substantially higher than anesthetic-related death rates in humans, which range from 0.02% to 0.05%. The difference in anesthetic-related death rates between animals and humans has been attributed, in large part, to a lack of monitoring or monitoring by inexperienced personnel.²⁴ Risk factors also play a part in anesthetic-related death rates, and certain breeds (ie, brachycephalic breeds) tend to have a higher risk than others.²⁴

Recognition of anesthetic-related complications in the perioperative period is important in decreasing the risk of anesthetic-related death. A study²⁵ performed at a university-based veterinary teaching hospital found a 0.005% incidence of cardiac arrest necessitating cardiopulmonary cerebral resuscitation in anesthetized patients. Of those anesthetized patients receiving cardiopulmonary cerebral resuscitation, 47% were resuscitated successfully.²⁵ Conversely, only 4% of non-anesthetized dogs undergoing cardiopulmonary arrest in an intensive care setting reportedly survive to discharge.²⁶ This difference is likely due to the higher incidence of concurrent diseases among patients in intensive care settings. According to the confidential enquiry into perioperative small animal fatalities, the 48-hour postoperative period has the highest risk of death, with almost 50% of anesthetic-related deaths in dogs occurring during this time.^23,24 Therefore, monitoring during the postoperative period is just as important as monitoring during the intraoperative period and should not be ignored.

Although the dog described in the present report was successfully resuscitated, there were indisputably other ways to approach anesthetic management of this case and management at the time of arrest. In critically reviewing the case, we found several controversial areas that deserve further explanation. First, the patient had hypocapnia and likely hypothermia during MRI, with a Petco between 25 and 29 mm Hg and a body temperature at the start of surgery of 34.5°C (94.1°F). Ideally, this patient's Petco₂ should have been maintained between 35 and 45 mm Hg.¹ Unfortunately, the MRI-compatible ventilator that was used is difficult to adjust for small patients and minor changes in peak inspiratory pressure or RR are problematic. Taking into account the existence of a gradient and the idea that Paco₂ is typically approximately 5 mm Hg higher than measured Petco₂,²⁷ the patient's Paco₂ was probably close to 30 to 34 mm Hg during MRI. This does not negate the fact that this patient was mildly over ventilated, but it probably was not overventilated to such a degree as to cause substantial metabolic alterations. Also, hypothermia possibly exacerbated the patient's bradycardia and hypotension during MRI. The patient required an anticholinergic to treat hypotension because hypothermia causes a decrease in cardiac output along with depression of the sinoatrial node and bundle of His.⁷ It is also possible that the patient's body temperature did not allow it to respond appropriately to anticholinergic administration.⁷ Ideally, more frequent monitoring of body temperature would have allowed for more aggressive warming of the patient in a more timely fashion. However, the monitoring unit in our MRI suite did not allow for body temperature monitoring; therefore, the patient could not be actively warmed until it was moved into the surgical suite. Second, the dose of atropine chosen was 0.02 mg/kg. This choice was questionable because some authors have suggested that too low a dose of atropine may aggravate vagally mediated bradycardia.^12,28 However, some authors suggest a lower dose of atropine be used if arrest is thought to be vagally induced, and suggested doses range from 0.02 to 0.04 mg/kg (0.009 to 0.018 mg/lb), IV.²⁹ In human medicine, atropine has even been removed entirely from some arrest treatment algorithms because its benefit in the treatment of pulseless electrical activity or asystole has been questioned in recent years.³⁰ Third, a high dose of epinephrine was used in the dog described in the present report, but the use of a high versus a low dose of epinephrine is controversial. Some authors consider a low dose of epinephrine to be more beneficial in the treatment of cardiac arrest because it is less arrythmogenic and causes less tachycardia. Currently, low-dose epinephrine treatment is the gold standard treatment in human medicine.²⁶ Yet, debate remains in veterinary medicine, as some authors recommend high-dose epinephrine treatment as more appropriate for cardiopulmonary cerebral resuscitation in dogs.^12,29 Fourth, the decision to administer sodium bicarbonate during the second arrest in the dog described in the present report is debatable, especially given that results of a current blood gas analysis were not available to evaluate the acid-base status of the patient prior to administration. Because this patient had been clinically dead for such a prolonged period (10 minutes) during the first arrest, it was likely that it had acidemia that predisposed it to the second arrest, and a clinical judgement was made that the dog would benefit from sodium bicarbonate administration.²⁹ Ideally, blood gas analysis should have been performed to determine whether sodium bicarbonate administration was indicated. Fifth, dobutamine and dopamine were used concurrently in this patient following the discontinuation of the epinephrine CRI. Use of dobutamine following resuscitation from cardiac arrest is not uncommon because dobutamine is helpful in normovolemic patients that need inotropic support.² We elected to administer dopamine also because of the potential that it could stimulate dopamine receptors in the renal circulation and stimulate additional β-adrenergic receptors.² A previous study³¹ that involved dogs anesthetized with isoflurane did not show a benefit to the use of dopamine and dobutamine together; however, that study was limited in the dosages used. In the authors' experience, the simultaneous use of dopamine and dobutamine can be beneficial in certain situations.

In the dog described in the present report, the sudden cardiopulmonary arrest was likely due to a chain of events that encouraged an overwhelming vagal response and led to bradycardia and asystole. These events were suspected to be aggravated by breed type, opioid administration, and extubation and tracheal manipulation. In addition, timing of otherwise innocuous actions may have compounded the physiologic response. This patient was successfully resuscitated because of the prompt initiation of appropriate treatment. Whether 1 or all 3 theoretical causes for the cardiac arrest in this patient were factors, it is difficult to assume that altering the course of events would have made a difference. It is important to monitor for and recognize any and all possible adverse reactions a patient may have during the course of anesthesia and recovery.