History

An 18-year-old 561.0-kg (1,234.2-lb) American Quarter Horse mare was referred to the University of Florida Large Animal Hospital because of signs of colic for > 7 hours. The referring veterinarian treated the mare with flunixin meglumine (1.1 mg/kg [0.5 mg/lb], IV) about 6 hours before the referral examination. The mare's previous medical history was unremarkable. On referral examination, the mare appeared reluctant to walk, was sweating and trembling, and had a varying heart rate (HR) of 64 to 82 beats/min (reference range, 30 to 45 beats/min), respiratory rate (RR) of 24 breaths/min (reference range, 8 to 20 breaths/min), pale pink to gray mucous membranes, capillary refill time of > 3 seconds, and weak but synchronous pulses.

Abdominal ultrasonography^a revealed gas distention of the large colon. A complete rectal examination was prevented by the presence of the gas-distended large colon in the pelvic canal. Results of plasma biochemical analyses indicated marked hypertriglyceridemia (581 mg/dL; reference range, 4 to 44 mg/dL), hyperglycemia (581 mg/dL; reference range, 71 to 122 mg/dL), hyperlactatemia (17 mmol/L; reference limit, < 2 mmol/L); moderate hyponatremia (126.1 mEq/L; reference range, 134 to 142 mEq/L); mild hypochloremia (noncorrected and corrected chloride concentrations, 79.7 mEq/L and 88.4 mEq/L, respectively; reference range, 95 to 104 mEq/L); and high creatinine concentration (3.59 mg/dL; reference range 0.8 to 1.5 mg/dL). However, Hct and plasma concentrations of BUN, potassium, calcium, and total protein were within reference limits.

Large colon torsion was suspected, and in preparation for surgical treatment, a 14-gauge 5.5-inch catheter was aseptically placed in the left jugular vein, through which fluid therapy (isotonic solution^b of balanced electrolytes [5 L, IV] and 7.2% saline [NaCl] solution [2 L, IV]) was administered over approximately 45 minutes. Afterward, the mare had unchanged HR and RR but mild improvement in the mucous membrane color and capillary refill time. The mare was sedated with detomidine hydrochloride (10.7 μg/kg [4.9 μg/lb], IV) and butorphanol tartrate (17.8 μg/kg [8.1 μg/lb], IV). General anesthesia was induced with ketamine hydrochloride (2.9 mg/kg [1.3 mg/lb], IV) and diazepam (0.05 mg/kg [0.02 mg/lb], IV) and maintained with isoflurane (2% to 2.5% initial vaporizer setting) administered in oxygen through a circle system.^c Immediately after intubation, intermittent volume-controlled, pressure-cycled positive-pressure ventilation^d was initiated (RR, 6 breaths/min; tidal volume, 7 L; peak inspiratory pressure, 40 cm H₂O; and positive end-expiratory pressure, 10 cm H₂O). Administration of a multielectrolyte solution^e (6.0 mL/kg/h [2.7 mL/lb/h], IV) was initiated, and the mare was instrumented for monitoring^f with 3-lead ECG, capnography, and pulse oximetry. Peripheral pulses were assessed, and because they could not be palpated, dobutamine (0.5 μg/kg/min [0.2 μg/lb/min], IV constant rate infusion [CRI]) was administered. In addition, a 20-gauge, 1.25-inch catheter was aseptically placed in the left facial artery for use in monitoring direct arterial blood pressure. The initial measurements of systolic, diastolic, and mean direct arterial pressures indicated hypertension (160, 80, and 110 mm Hg, respectively; reference ranges, 90 to 130, 60 to 90, and 70 to 110 mm Hg, respectively). Hypertension was attributed to the cardiovascular effects of the dobutamine CRI combined with previously administered detomidine^1,2; thus, the dobutamine CRI was reduced to 0.30 μg/kg/min (0.14 μg/lb/min). Afterward, the systolic, diastolic, and mean direct arterial pressures decreased to 110, 58, and 81 mm Hg, respectively. The anesthetic plane was deemed appropriate, and surgery began. Lidocaine (1 mg/kg [0.45 mg/lb], IV bolus, then 50.0 μg/kg/min [22.7 μg/lb/min], IV CRI) was administered for its antinociceptive, volatile anesthetic-sparing, prokinetic, and anti-inflammatory effects.³ Butorphanol (0.010 mg/kg [0.005 mg/lb], IV) was also administered; however, its antinociceptive effects are questionable.⁴ Gentamicin sulfate (6.6 mg/kg [3.0 mg/lb], IV over 30 minutes) was administered prophylactically for its broad-spectrum antimicrobial and bactericidal effects.

Throughout the first hour of anesthesia, rapid changes occurred multiple times in arterial pressure (mean arterial pressure [MAP] range, 40 to 100 mm Hg) and HR (range, 60 to 80 beats/min). Results of initial arterial blood gas and biochemical analyses^g (40 minutes after induction) indicated metabolic acidosis, mild hypoxemia, and hyperglycemia (Table 1); therefore, albuterol (0.54 mg) was administered endotracheally. Subsequent analyses performed during anesthesia revealed improved oxygenation but worsened lactic metabolic acidosis.

One hour after induction, the mare had sporadic triplets of ventricular premature complexes that progressed to runs of monomorphic ventricular tachycardia and spontaneously converted to sinus tachycardia without treatment. In addition, HR and blood pressure varied, with severe tachycardia (HR, 118 beats/min) and hypertension (MAP, 140 mm Hg) followed by moderate tachycardia (HR, 65 beats/min) and hypotension (MAP, 50 mm Hg) that seemed unrelated to surgical stimulation or dobutamine administration. A catheter was aseptically placed in the right jugular vein, through which additional CRIs of a multielectrolyte solution^e (12.0 mL/kg/h [5.5 mL/lb/h]) and hydroxyethyl starch (2.4 mL/kg/h [1.1 mL/lb/h]) were initiated.

The mare's MAP suddenly dropped to 25 mm Hg; thus, ephedrine sulfate (0.05 mg/kg, IV) was administered and resulted in transient hypertension (MAP, 135 mm Hg) and sinus tachycardia (HR, 180 beats/min) that lasted approximately 3 minutes. However, sinus tachycardia (HR, 120 to 100 beats/min) and hypotension (MAP, 35 to 40 mm Hg) followed and persisted. Five boluses of phenylephrine hydrochloride (each 1.8 to 3.5 μg/kg [0.8 to 1.6 μg/lb], IV) were administered but did not change the mare's blood pressure or HR. A blood sample was collected and submitted for measurement of cardiac troponin I concentration, the lidocaine CRI was discontinued to avoid possible further vasodilation,^5,6 the dobutamine CRI was discontinued to limit its chronotropic effect, a CRI of norepinephrine bitartrate (0.05 μg/kg/min, IV initially, then progressively increased to 0.20 μg/kg/min [0.09 μg/lb/min], IV) was initiated, and the vaporizer setting for isoflurane was gradually reduced from 2% to 1%. Afterward, the mare had hemodynamic improvement (HR range, 60 to 80 beats/min; MAP range, 54 to 75 mm Hg) but continued metabolic acidosis that progressively worsened (Table 1).

The exploratory laparotomy was completed 180 minutes after induction of general anesthesia. The mare did not have colonic torsion, as suspected, but had impactions of the large colon, cecum, and small colon that were resolved with enterotomy and typhlotomy. The mare was then moved to a recovery stall, and positive-pressure ventilation with 100% oxygen was provided with a Hudson demand valve^h until spontaneous ventilation resumed. Supplemental oxygen (10 L/min) was then supplied through a urinary catheter connected to an oxygen flowmeter with humidifierⁱ and inserted into the mare's endotracheal tube. Approximately 45 minutes after discontinuation of isoflurane, the mare swallowed repeatedly and was then extubated. Supplemental oxygen and the CRIs of multielectrolyte solution^e and norepinephrine were also discontinued at that point. Overall, 23 L of multielectrolyte solution^e and 2.5 L of hydroxyethyl starch had been administered during surgery and recovery.

The mare made attempts to move while still having fast nystagmus. Therefore, butorphanol (10 μg/kg [4.5 μg/lb], IV) and 3 boluses of xylazine (0.18 mg/kg [0.08 mg/lb], 0.09 mg/kg [0.04 mg/lb], and 0.18 mg/kg, IV) were administered for analgesia and sedation during recovery.

Approximately 110 minutes after being moved to a recovery stall, the mare stood up well, but soon developed muscle fasciculations and profuse sweating that spontaneously resolved within a few minutes. Approximately 30 minutes after standing, the mare appeared quiet and well coordinated and was therefore walked to an assigned stall.

Approximately 10 hours after recovery, the mare began to have episodes of sinus tachycardia (HR, 80 to 88 beats/min) and profuse sweating. About an hour later, the mare developed signs of shock, with severe tachycardia (HR range, 100 to 120 beats/min), weak pulses, dark-pink mucous membranes, prolonged capillary refill time (> 3 seconds), labored and stertorous respirations, tachypnea (RR, 30 breaths/min), muscle fasciculations, swaying, postural instability, and extremities that felt cold to the touch. Hematologic results indicated disseminated intravascular coagulopathy, with 45,000 platelets/μL (reference range, 94,000 to 232,000 platelets/μL). The mare's Hct was 26%; plasma concentrations of total protein and creatinine were 3.3 g/dL and 6.5 mg/dL, respectively; and plasma lactate concentration was too high to be measured. The mare developed incessant multifocal ventricular tachycardia (HR, 112 beats/min), spontaneous hemorrhage from the surgical incision, and epistaxis.

Question

What could have caused the severe cardiovascular instability in this mare during and after general anesthesia?

Answer

The mare's hemodynamic instability may have been caused by catecholamine release from a pheochromocytoma.

The owner elected euthanasia for the mare. A full necropsy was performed approximately 24 hours later and revealed evidence of massive hemorrhage in the retroperitoneal space, extending from the level of the diaphragm to the subperitoneal pelvic space (Figure 1). The right adrenal gland was ruptured and effaced by an expansile mass, necrosis, and semicoagulated blood. Histologic examination of samples from the right adrenal gland revealed extensive effacement of the adrenal architecture by packets and nests of neoplastic cells, each with granular cytoplasm, round nucleus, and a single nucleolus. There was evidence of hemorrhage and necrosis throughout the neoplasm and invasion of neoplastic cells into tissues surrounding the right adrenal gland. These neoplastic cells had positive results for immunohistochemical staining for cytoplasmic synaptophysin,^j a confirmatory marker for pheochromocytoma. Additionally, the mare had pericardial and peritoneal serosanguineous effusion, renal hemorrhagic foci, and multifocal areas of petechial hemorrhage, findings consistent with disseminated intravascular coagulation. Histologic examination of heart tissue revealed occasional degenerative and necrotic cardiac myocytes.

View Full Size

Postmortem images of the pheochromocytoma, in situ (A) and isolated (B), identified on necropsy of an 18-year-old 561.0-kg (1,234.2-lb) American Quarter Horse mare that underwent surgical treatment for colic. A—Semiclotted blood fills the retroperitoneal space (asterisk) adjacent to the diaphragm (long arrow), and multifocal petechial hemorrhages (short arrows) are evident, consistent with disseminated intravascular coagulation. The mare's head is toward the right of the image. B—The right adrenal gland is ruptured and effaced by a dark red expansile mass, with only remnant cortex (arrowheads) evident.

Citation: Journal of the American Veterinary Medical Association 257, 2; 10.2460/javma.257.2.151

• Download Figure
• Download figure as PowerPoint slide

Results for cardiac troponin I concentration analysis submitted intraoperatively were markedly high (6.49 ng/mL; reference range, < 0.1 ng/mL⁷), indicative of severe myocardial injury. A sample of the mare's blood retained from those collected during anesthesia was submitted postmortem for measurement of the metanephrine-to-creatinine and normetanephrine-to-creatinine ratios, along with a submitted comparative blood sample collected from a control horse (a healthy gelding selected from the pool of teaching-research horses of the University of Florida). Results indicated that the mare's metanephrine-to-creatinine and normetanephrine-to-creatinine ratios were 12 and 34 times those for the control horse, respectively (reference limit, < 4 times those for a control horse⁸), consistent with the postmortem diagnosis of pheochromocytoma.

Discussion

Pheochromocytoma is a rare catecholamine-secreting tumor derived from chromaffin cells of the adrenal medulla. A retrospective study⁹ of 4,094 horses necropsied shows that 37 (0.9%) horses had pheochromocytoma, the tumor was often (30/37 [81%]) an incidental finding, and most horses affected were older (age range, 13 to 38 years). In addition, pheochromocytoma does not seem to have a breed or sex predilection in horses.^9,10

Clinical signs of pheochromocytoma in horses may include tachycardia, tachypnea, profuse sweating, excitation, and muscle fasciculations, signs that are nonspecific and that can occur with conditions such as abdominal pain and endotoxemia^9,10; therefore, pheochromocytoma was not initially suspected during anesthesia. However, the mare's severe intraoperative cardiovascular instability did not seem to correspond with surgical events, and fluid therapy yielded minimal improvement in cardiovascular status; thus, we knew that nociception and hypovolemia were not the only causes of the mare's cardiovascular instability.

Ventricular premature complexes and runs of ventricular tachycardia occurred in the mare of the present report during anesthesia. A lidocaine CRI was used to provide additional analgesia during general anesthesia and for its prokinetic and anti-inflammatory properties³; therefore, no additional intraoperative treatments were administered for the runs of ventricular tachycardia.

Ventricular arrhythmias are not unique signs of pheochromocytoma in horses. Arrhythmias in horses with gastrointestinal disease may be caused by hypovolemia, endotoxins, or autonomic imbalance from gastrointestinal distension or metabolic, electrolyte, or acid-base imbalances.¹¹

Hyperglycemia, hypertriglyceridemia, and hyperlactatemia detected in the mare of the present report could have been caused by gastrointestinal disease^12,13 or the pheochromocytoma^9,10 that was identified postmortem. In humans, hypercatecholaminemia stimulates lipolysis, which leads to marked hyperglycemia and hypertriglyceridemia, whereas severe catecholamine-induced vasoconstrictive ischemia may lead to lactic acidosis.¹⁴ In the mare of the present report, blood lactate concentrations did not decrease as expected with fluid therapy. Sodium bicarbonate was not supplemented because its use is not supported in the context of lactic acidosis.¹⁵ In fact, direct manipulation of blood pH has not been shown to improve cardiovascular function in the context of high anion gap metabolic acidosis, and the increase in carbon dioxide caused by bicarbonate administration may be associated with paradoxical intracellular acidosis.¹⁵ Although hyperlactatemia could have been a consequence of hemodynamic instability, neither condition was explained by surgical findings, which did not include colonic torsion or intestinal hypoperfusion and therefore suggested a cause other than gastrointestinal disease.

Consistent with findings of a retrospective study⁹ of horses with pheochromocytoma, the mare of the present report had severe sweating and muscle fasciculations. These signs could have also been associated with administration of α₂-adrenoceptor agonists¹ or with pain and stress resulting from colic.¹² In addition, hypocalcemia may have contributed to the mare's muscle fasciculations. Although the mare had hyponatremia on presentation, results of subsequent biochemical analyses indicated that the condition had been corrected with fluid therapy.

The finding of hypertension in the mare of the present report was consistent with findings in dogs¹⁶ and humans¹⁴ with pheochromocytomas; however, to our knowledge, no studies have focused on blood pressure monitoring in horses with pheochromocytomas. A recent report¹⁷ described monitoring oscillometric blood pressure measurements in a horse that underwent general anesthesia and that was later determined to have a pheochromocytoma; however, that horse did not have hypertension during the anesthetic event. Hypertension in the mare of the present report was initially attributed to the vasoconstrictive effect of detomidine¹ in combination with the positive inotropic effect of dobutamine.² Other possible causes could have included inadequate anesthetic depth, inadequate analgesia, hypoxemia, and hypercapnia¹⁸; however, the anesthetic depth was deemed appropriate at all times, the episodes of hypertension did not appear related to surgical stimulation, and analgesics were administered intraoperatively. Although hypoxemia (Pao₂, 76 mm Hg) was detected on the first blood gas analyses, subsequent blood gas analyses revealed Pao₂ > 100 mm Hg, with clinically normal Paco₂, making hypoxemia and hypercapnia unlikely causes of hypertension in this case. An accidental overdose of a sympathomimetic drug also could have caused hypertension in this mare; however, such was ruled out during the hypertensive episodes, and the dobutamine CRI was adjusted in response to the changes in the mare's blood pressure. Further, the mare's response to dobutamine and ephedrine seemed excessive and hard to titrate and control.

In horses, presumptive antemortem diagnosis of a pheochromocytoma can be challenging because the clinical signs can be vague, nonspecific, and intermittent, and definitive diagnosis of pheochromocytomas in horses can only be confirmed postmortem.^9,10 In contrast, definitive antemortem diagnosis of functional pheochromocytoma can be made in humans¹⁹ and dogs²⁰ because in these species, plasma-free metanephrines and urinary fractionated metanephrine concentrations are associated with the presence of catecholamine-secreting tumors. However, these tests may be less specific for a diagnosis of pheochromocytoma in horses because other conditions, including colic and laminitis, may also cause increased metanephrine concentration as a result of catecholamine secretion in response to pain and stress.²¹ Nonetheless, pheochromocytoma should be included among differential diagnoses in horses with a plasma metanephrine-to-creatinine ratio > 4 times the ratio in a control horse,⁸ as was the case in the mare of the present report. Although clinical signs of pheochromocytoma in horses are nonspecific and may be observed in horses with gastrointestinal disease, the findings in the mare of the present report highlighted the importance of considering pheochromocytoma as a differential diagnosis in horses with signs of abdominal pain combined with hyperglycemia, hyperlactatemia, sweating, tremors, and uncontrollable hemodynamic instability characterized by tachycardia, ventricular arrythmias, erratic response to sympathomimetic administration, and episodes of hyper- and hypotension during anesthesia.