History

A 7.5-year-old 27.5-kg (60.5-lb) castrated male Boxer was referred to the Washington State University Veterinary Teaching Hospital for evaluation and treatment of an abdominal mass. The patient had a 1-month history of exercise intolerance and had developed swelling of the left tarsal joint that resolved following administration of amoxicillin and prednisolone prescribed by the referring veterinarian. The owner reported that the dog had become incontinent and leaked urine when lying down or shifting position. The dog also had hematuria, but the owner reported no straining during urination or defecation, no change in frequency of urination, and no vomiting or diarrhea. Abdominal radiography performed by the referring veterinarian revealed a large mass in the caudodorsal aspect of the abdomen that was impinging on the bladder. The patient was being treated with thyroxine (0.5 mg, PO, q 12 h) because of hypothyroidism, which had been diagnosed approximately 3 years earlier.

On initial examination at the Veterinary Teaching Hospital, the dog was bright, alert, and responsive. Body condition score was 4 on a scale from 1 to 5. Oral mucous membranes were pink and moist, and capillary refill time was < 2 seconds. Auscultation of the heart and lungs did not reveal any abnormalities. Abdominal palpation revealed a large, firm mass in the caudodorsal aspect of the abdomen that was approximately 10 to 15 cm in diameter. The remainder of the physical examination was unremarkable.

A CBC and serum biochemistry profile were performed. Abnormalities included neutrophilia (15,664 neutrophils/ML; reference range, 2,300 to 8,600 neutrophils/ML), lymphopenia (890 lymphocytes/ML; reference range, 1,500 to 7,700 lymphocytes/ML), thrombocytosis (705,000 platelets/ML; reference range, 157,000 to 394,000 platelets/ML), hypercholesterolemia (432 mg/dL; reference range, 135 to 278 mg/dL), and slightly high alkaline phosphatase activity (111 U/L; reference range, 14 to 72 U/L). One-stage prothrombin time was slightly prolonged (9.9 seconds; reference range, 6.4 to 8.2 seconds), but activated partial thromboplastin time was within reference limits (10.8 seconds; reference range, 8.4 to 14.8 seconds), and buccal mucosal bleeding time was 85 seconds (reference range, < 4 minutes). Urine specific gravity was 1.014, and dipstick analysis of a urine sample revealed 2+ protein, 2+ hemoglobin, and 1+ bilirubin. Bacterial culture of a urine sample did not yield any growth. The urine protein-to-creatinine concentration ratio was 4.4 (reference range, < 1).

Thoracic radiography did not reveal any abnormalities. Abdominal ultrasonography revealed a large heterogeneous and complex mass occupying the left caudal portion of the abdomen. Cytologic examination of a fine-needle aspirate of the abdominal mass revealed mild suppurative inflammation and evidence of necrosis; findings were considered suggestive of a neoplastic process.

Abdominal exploratory surgery and removal of the mass were planned. Prior to surgery, the dog was assigned an American Society of Anesthesiologists status of III on the basis of the underlying disease process. The dog was premedicated with hydromorphone (0.2 mg/kg [0.09 mg/lb], IM), and 30 minutes later, 18-gauge indwelling catheters were placed in the left cephalic vein and right saphenous vein. Electrocardiography leads were placed for assessment of heart rhythm, and a size 4 blood pressure cuff was placed on the right forelimb for indirect measurement of blood pressure. The dog was preoxygenated with 100% oxygen delivered through a face mask for approximately 5 minutes, and anesthesia was induced with propofol administered to effect (administered dose, 1.5 mg/kg [0.68 mg/lb], IV). A 10-mm endotracheal tube was placed in the trachea by means of direct visualization with a laryngoscope. Anesthesia was maintained with isoflurane in oxygen (flow rate, 2 L/min) delivered with an out-of-circle precision vaporizer through a circle breathing system. A pulse oximeter probe was placed on the tongue, and a 20-gauge catheter was placed in the right dorsal pedal artery and connected to a disposable calibrated pressure transducer^a for direct measurement of arterial blood pressure. While the patient was being prepared for surgery, the lumbosacral area was clipped free of hair and surgically scrubbed. Preservative-free morphine (0.1 mg/kg [0.045 mg/lb]) added to isotonic saline (0.9% NaCl) solution to obtain a final volume of 6 mL was injected into the epidural space at the L7-S1 junction.

In the operating room, the dog was positioned in dorsal recumbency and connected to a multipurpose monitor^b that recorded the ECG, oxygen saturation as measured by pulse oximetry, indirect blood pressure measurements, end-tidal partial pressure of carbon dioxide, core body temperature, and direct arterial blood pressure measurements. Rectal temperature was maintained between 36.7° and 37.8°C (98.0° and 100.0°F) with the help of a convective air warming system,^c warm water blankets,^d and a fluid line warming system.^e Lactated Ringer's solution was administered at a rate of 10 mL/kg/h (4.5 mL/lb/h), IV. Throughout the procedure, systolic, diastolic, and mean arterial blood pressures remained between 90 and 140 mm Hg, 40 and 65 mm Hg, and 55 and 85 mm Hg, respectively. Heart rate ranged from 60 to 110 beats/min.

Approximately 1 hour after surgery was begun, an arterial blood sample was obtained and submitted for blood gas analysis^f; pH was 7.31, arterial partial pressure of CO₂ (PaCO₂) was 45 mm Hg, arterial partial pressure of oxygen (PaO₂) was 444 mm Hg, bicarbonate concentration was 22.8 mEq/L, sodium concentration was 149 mEq/L, potassium concentration was 4.6 mEq/L, and base excess was −2.8 mEq/L. Approximately 2 hours after surgery was begun, the PCV was 25%, compared with a preoperative value of 39%. The decrease in PCV was attributed to a combination of intraoperative blood loss and crystalloid fluid administration. A blood transfusion was started, and the lactated Ringer's solution was changed to a balanced electrolyte solution^g to avoid any possibility that calcium in the lactated Ringer's solution might promote clotting of the transfused blood. An arterial blood sample was submitted for blood gas analysis 1 hour after the blood transfusion was begun (3 hours after surgery was begun), and pH was 7.23, PaCO₂ was 66 mm Hg, PaO₂ was 380 mm Hg, bicarbonate concentration was 28.0 mEq/L, sodium concentration was 148 mEq/L, potassium concentration was 5.8 mEq/L, and base excess was 0.4 mEq/L. Mechanical ventilation was instituted at a rate of 11 breaths/min with a tidal volume of 400 mL. Thirty minutes later, analysis of a follow-up arterial blood sample revealed a pH of 7.39, PaCO₂ of 34 mm Hg, PaO₂ of 481 mm Hg, bicarbonate concentration of 20.7 mEq/L, sodium concentration of 146 mEq/L, potassium concentration of 7.0 mEq/L, and base excess of −3.1 mEq/L. Despite the hyperkalemia, no change was seen in the morphology of the ECG waveform.

The blood transfusion was discontinued, and infusion of a 5% dextrose solution was initiated in an attempt to move extracellular potassium to the intracellular compartment.¹ After surgery, the dog was admitted to the intensive care unit and infusion of the 5% dextrose solution was continued for approximately 1 hour, after which time the serum potassium concentration had decreased to 4.4 mEq/L. The patient recovered without further complications and was discharged from the hospital 4 days later. Results of histologic examination of the abdominal mass were consistent with a diagnosis of hemangiosarcoma.

Question

What was the cause of the intraoperative hyperkalemia in this patient?

Answer

The most likely cause of hyperkalemia in this patient was acute intraoperative tumor lysis.

Discussion

Potential causes of hyperkalemia in this dog include thrombocytosis, respiratory acidosis, intravascular hemolysis, and intraoperative tumor lysis.

Thrombocytosis can be seen in conjunction with various neoplastic processes, bone marrow stimulation secondary to blood loss, and various endocrine disorders, such as hypothyroidism and hyperadrenocorticism, and has previously been reported to cause hyperkalemia.² However, serum potassium concentration in the dog described in the present report was within reference limits prior to surgery (5.2 mEq/L; reference range, 4.4 to 5.3 mEq/L), excluding these conditions as preexisting causes of hyperkalemia in this dog. In addition, the hyperkalemia associated with thrombocytosis is usually mild.²

Acidosis can result in hyperkalemia by causing extracellular movement of potassium ions in exchange for intracellular movement of hydrogen ions.² The dog described in the present report did develop mild respiratory acidosis while anesthetized, which is common in anesthetized patients. However, even as mechanical ventilation reduced the PaCO₂ from 66 to 34 mm Hg and increased the pH from 7.23 to 7.39, serum potassium concentration increased from 5.8 to 7.0 mEq/L, ruling out respiratory acidosis as a cause of hyperkalemia in this patient.

Intravascular hemolysis can cause hyperkalemia secondary to release of intracellular potassium.² Thus, when hyperkalemia was detected in the dog described in the present report, the blood transfusion was discontinued until hemolysis could be ruled out as the underlying cause. No macroscopic or microscopic evidence of hemolysis was found, suggesting that intravascular hemolysis was unlikely to be the cause of the hyperkalemia in this dog. In addition, although English Springer Spaniels and some Japanese and East Asian dog breeds, such as the Akita, Shinshu Shiba, and San'in Shiba, have been shown to have higher RBC potassium concentrations than dogs of other breeds,^2,3 the blood transfusion administered in this case had been obtained from a Labrador Retriever, which would not be expected to have RBCs with high potassium content. The supernatant of stored units of human RBCs has been reported to have potassium concentrations > 60 mEq/L,⁴ but the potassium concentration of the transfused blood administered to the dog described in the present report was only 4 mEq/L, which is in the acceptable range for stored blood. Finally, hyperkalemia related to transfusion has been shown to depend not only on the potassium concentration of the donor RBCs but also on volume and rate of RBC administration,⁵ and in the present case, the rate of blood transfusion and the amount of blood administered would not have been expected to cause the observed increase in serum potassium concentration.

Acute tumor lysis associated with tumor manipulation or handling during surgery has been described in humans^6–8 and can lead to complications such as hyperkalemia, hyperphosphatemia, hyperuricemia, and hypocalcemia.⁹ Hyperkalemia is the most immediate and most dangerous consequence of acute tumor lysis and may be exacerbated by hypocalcemia, acidosis, or renal failure. Given that other causes of hyperkalemia were ruled out in the dog described in the present report, the most likely cause was acute tumor lysis.

The pathogenesis of tumor lysis associated with tumor manipulation or handling during surgery is unclear.^10,11 Several clinical factors are associated with an increased risk of acute tumor lysis syndrome in humans, including a large tumor burden, high proliferative fraction, extreme sensitivity to chemotherapeutic drugs, abdominal involvement, and high baseline lactate dehydrogenase and uric acid concentrations.^12,13 The tumor in the dog described in the present report had attachments to the iliopsoas muscle, abdominal aorta, and caudal vena cava, and the ureter was adhered to the capsule of the mass dorsomedially. In addition, there were multiple branches from the aorta that penetrated the mass. The abdominal involvement and invasive and infiltrative nature of the tumor might have predisposed it to release substances from necrosing tumor cells or leak hemorrhagic fluid into the systemic circulation during handling, which could have led to hyperkalemia.

Acute tumor lysis syndrome is a common complication in human patients with cancer, with the highest incidence in patients with hematologic malignancies,¹⁴ although it can also be seen in patients with solid tumors.⁹ Immunotherapy, chemotherapy, radiation therapy, and corticosteroid administration can precipitate acute tumor lysis syndrome,^15–18 although none of these were a factor in the dog described in the present report, other than administration of prednisolone prescribed by the referring veterinarian to treat swelling of the tarsal joint. Acute tumor lysis syndrome has also been reported to be associated with anesthesia.¹⁹

With acute tumor lysis syndrome, intracellular contents, such as phosphorus, potassium, and purines, are released into the systemic circulation as tumor cells are destroyed. If the concentration of these substances exceeds the excretory capacity of the kidneys, lifethreatening metabolic and electrolytic abnormalities may occur. Although purines in humans are catabolized by the liver through oxidation of xanthine and hypoxanthine to uric acid,²⁰ in dogs other than English Bulldogs and Dalmatians, uric acid is oxidized to allantoin by uricase,^21–23 with the result that most dogs would not be expected to develop hyperuricemia. Phosphate released from lysed cells can combine with calcium to form calcium phosphate, an insoluble compound that may lead to tissue damage and cause renal failure, pruritus, cutaneous gangrene, and inflammation of the eyes or joints,¹⁹ although none of these abnormalities were identified in the dog described in the present report.

Clearance by the kidneys is the main mechanism by which phosphate, potassium, and uric acid are excreted. Thus, preexisting renal disease or dehydration can exacerbate the effects of acute tumor lysis syndrome.¹² Preoperative urinalysis in the dog described in the present report revealed hyposthenuria, a high protein content, and a high urine protein-to-creatinine ratio, which could be suggestive of glomerulopathy. However, SUN and creatinine concentrations were within reference limits, and the dog was producing normal amounts of urine.

Treatment of acute tumor lysis syndrome is mainly supportive and is directed at maintaining hydration and controlling hyperkalemia.^21–23 Appropriate treatment of hyperkalemia depends on the rapidity of onset and the magnitude of the increase in potassium concentration. Changes in the ECG waveform associated with hyperkalemia were described in a study²⁴ in which hyperkalemia was induced in healthy animals. In that study, serum potassium concentrations of 5.5 to 6.5 mEq/L were associated with increases in T wave amplitude; concentrations of 6.6 to 7.0 mEq/L were associated with decreases in ≥ wave amplitude, widening of the QRS complex, prolongation of the PR interval, and depression of the ST segment; concentrations of 7.1 to 8.5 mEq/L were associated with decreases in P wave amplitude, increases in P wave duration, and prolongation of the QT interval; concentrations of 8.6 to 10.0 mEq/L were associated with a lack of P waves (atrial standstill) and sinoventricular rhythm; and concentrations > 10.1 mEq/L were associated with widening of the QRS complex, eventual replacement of the QRS complex with a smooth biphasic waveform, ventricular flutter, ventricular fibrillation, and asystole. However, these characteristic changes may not always be apparent clinically^25,26 because concurrent abnormalities in serum sodium, chloride, magnesium, and calcium concentrations and venous pH in clinically ill patients with hyperkalemia may have their own effects on the ECG waveform. In the dog described in the present report, no changes in the ECG waveform were observed.

Hyperkalemia should be treated regardless of the magnitude of the serum potassium concentration if ECG abnormalities are observed. Various strategies for treating hyperkalemia that have been described include administration of calcium gluconate to antagonize the effects of potassium on the cell membrane; administration of sodium bicarbonate or glucose, with or without concurrent administration of insulin, to induce movement of potassium from the extracellular fluid to the intracellular fluid; and removal of potassium from the body with cation exchange resin or dialysis.¹ Administration of calcium normalizes the difference between resting and threshold potential and restores membrane excitability. Administration of glucose increases endogenous insulin release, causing movement of potassium into cells. Administration of sodium bicarbonate causes movement of potassium ions into cells as hydrogen ions leave the cells to buffer the additional bicarbonate in the extracellular fluid.

To our knowledge, acute tumor lysis syndrome associated with surgical manipulation of a tumor in a dog has not been reported previously. Anesthesiologists should be aware of the potential risk of acute tumor lysis syndrome leading to hyperkalemia, and we recommend that serum potassium concentration be monitored in patients undergoing surgical manipulation of large, invasive tumors. Importantly, animals with hyperkalemia may have normal ECG waveforms.²⁴ Therefore, ECG monitoring is not a substitute for measuring serum potassium concentration. Although the hyperkalemia in the dog described in the present report was managed without complications, intractable hyperkalemia leading to cardiac arrest has been reported in human patients,^6,7 and timely recognition of hyperkalemia can provide more time for treatment.