Wide-complex tachycardia associated with severe hyperkalemia in three cats

Abstract

The well recognized cardiac effects of severe hyperkalemia include progressive rhythm and conduction disturbances such as bradycardia, spiked and narrow T waves, widening QRS complex, widening and flattening P wave, disappearance of the P wave, and cardiac arrest. Paradoxically, a heart rate greater than 200 beats/min may coexist with hyperkalemia in some cats. This report describes three cats with moderate to severe hyperkalemia and concurrent rapid heart rate. In each cat, the serum potassium (K⁺) concentration was ≥7.5 mEq/dl with a concurrent heart rate>200 beats/min. In each cat, nine-lead electrocardiograms demonstrate an absence of P waves and a wide-complex tachycardia. Hyperkalemia should be considered in the differential diagnosis when a feline electrocardiogram demonstrates a wide-complex tachycardia without identifiable P waves.

Hyperkalemia produces well-known characteristic scalar limb lead electrocardiogram (ECG) abnormalities in animals and humans (Surawicz 1967a, 1967b). The ECG abnormalities generally become more striking and progress with increasing hyperkalemia. The earliest characteristic ECG change is slowing of the heart rate and peaking of the T wave (Winkler et al 1938). This is followed by widening of the QRS complex and flattening, widening, and then disappearance of the P wave. Increasing hyperkalemia leads to further widening of the QRS complex and blending with the T wave in a sine-wave morphology. Ultimately, ventricular fibrillation or asystole results in cardiac arrest. The presence of a slow heart rate and the ECG features listed above are clues that suggest hyperkalemia when a serum potassium level is pending (Feldman and Nelson 1996).

When moderate to severe hyperkalemia (serum K⁺>8.5 mEq/dl) (Feldman and Nelson 1996) occurs in cats, it is often a result of acute failure to excrete potassium through the urine. Urethral obstruction (Parks 1975, Schaer 1977), anuric renal failure, and reperfusion of injured skeletal muscle following an aortic embolism (Laste and Harpster 1995) are examples of conditions causing severe hyperkalemia in cats. Iatrogenic causes are reported rarely (Dhein and Wardrop 1995). In most clinical and experimental reports involving cats, hyperkalemia has been reported to produce heart rates in the range of 90–190 beats/min (Parks 1975, Kittleson 1998). Some anecdotal reports and individual cases have made exception to the association of bradycardia with hyperkalemia in cats (Schaer 1977, Tilley 1992, Côté and Ettinger 2005). This paper describes three cats, two with anuria and one with aortic thromboembolism, in which moderate to severe hyperkalemia coexisted with tachycardia, defined here as a heart rate greater than 200 beats/min (Hamlin 1989, Côté et al 2004).

Case reports

Case 1

An 8-year-old, 16.5 lb (7.5 kg) castrated male domestic shorthair cat was presented for evaluation of a 24-h history of lethargy, weakness, anorexia and vomiting. On physical examination, temperature was 102.5°F, heart rate was 108 beats/min, and a grade II/VI systolic heart murmur was present. The respirations were comfortable and the pulse was synchronous and mildly hypokinetic. The cat was hospitalized and received lactated Ringer's solution (LRS) (125 ml, subcutaneously). Twelve hours after admission, the heart rate was 116 beats/min and a serum chemistry panel revealed severe azotemia, mild hypochloremia, and moderate to severe hyperkalemia (8.7 mEq/l; normal range 3.1–5.2 mEq/l) (Hitachi 911 Analyzer, Hitachi Corporation, Indianapolis, IN). Abdominal radiographs indicated numerous cystic calculi in an otherwise empty urinary bladder, left renomegaly and renal mineralization. Treatment with intravenous fluids (0.9% sodium chloride at 20 ml/h) and famotidine (Pepcid; Merck & Co) (1 mg/kg IV, q 12 hours) was initiated. After 24 h, mild tachypnea and dyspnea were noted. The heart rate had now increased to 176 beats/min from 116 beats/min and the respirations were 44 beats/min at rest. Diagnostic imaging suggested either left sided congestive heart failure or circulatory overload and concurrent bilateral hydronephrosis with possible ureteral rupture, uroperitoneum, and uroabdomen. A second serum potassium level remained elevated (9.2 mEq/l; normal range 3.1–5.2 mEq/l) (i-stat, Heska Corporation, Fort Collins, CO). An ECG revealed wide-complex tachycardia with no P waves (Fig 1). Despite treatment with calcium gluconate (Phoenix Laboratories) (1.86 mEq, IV, over 10 min), sodium bicarbonate (Phoenix Laboratories) (10 mEq, IV, over 20 min), and furosemide (Lasix; Hoechst) (2 mg/kg IV, q 4 hours, twice), the cat's condition continued to deteriorate and the owner requested that the cat be euthanased.

Fig 1.

Nine-lead electrocardiogram from case 1 demonstrating wide-complex tachycardia and absent P waves. The lead labeled V1 is rV2 and the lead labeled V3 is V4. Heart rate=210 beats/min. Serum potassium=9.2 mEq/l. Paper speed=25 mm/s. Standardization is 1 cm/1 mV. Cat is in right lateral recumbency.

Case 2

An 11-year-old, 15.8 lb (7.2 kg) spayed female Siamese cat presented in lateral recumbency with a history of controlled diabetes mellitus. On physical examination, the cat was dehydrated, and the heart rate was 220 beats/min with no murmur or gallop ausculted. There was right rear limb paralysis; the limb was cold, had no palpable femoral pulse and cut toe nails did not bleed. The serum chemistry revealed an elevated serum potassium level (K⁺=7.5 mEq/l; normal range 3.1–5.2 mEq/l). Thoracic and abdominal radiographs, echocardiogram and an abdominal ultrasound examination were within normal limits with the exception of obstructed blood flow in the distal aorta as determined using color-flow Doppler. The ECG at presentation demonstrated wide-complex tachycardia and the absence of P waves (Fig 2a). Based on these findings, a diagnosis of aortic thromboembolism with reperfusion and secondary hyperkalemia was made.

Fig 2.

(a) Nine-lead electrocardiogram from case 2 demonstrating wide-complex tachycardia and absent P waves. The lead labeled V1 is rV2 and the lead labeled V3 is V4. Heart rate=240 beats/min. Serum potassium=7.5 mEq/l. Paper speed=25 mm/s. Standardization is 2 cm/1 mV. Cat is in right lateral recumbency. (b) Leads I, II, and III from case 2 after 24 h of treatment. A P wave is present for each QRS complex. The rhythm is sinus tachycardia. Heart rate=260 beats/min. Serum potassium=4.4 mEq/l. Paper speed=25 mm/s. Standardization is 1 cm/1 mV. Cat is in right lateral recumbency.

Initial treatment included intravenous fluid therapy (0.9% NaCl+2.5% dextrose at 30 ml/h), heparin (heparin sodium, Elkins-Sinn Inc) (140 mg/kg, SQ, q 6 hours), and regular insulin (Eli Lilly) (0.2 units/kg body weight IM, q 3 hours if blood glucose was >220 mg/dl). Twenty-four hours after presentation, the serum potassium concentration had decreased (4.4 mEq/l; normal range 3.1–5.2 mEq/l). The ECG at this time showed sinus tachycardia and first degree AV block (Fig 2b). There was still no femoral pulse and no use of the right hind limb. As pulselessness and paralysis persisted 48 h after presentation, the owners elected for euthanasia due to the poor prognosis.

Case 3

An 8-year-old, 11.3 lb (5.1 kg) spayed female domestic longhair cat presented for evaluation of acute vomiting and diarrhea. The cat's recent history included chemotherapy for suspected mediastinal lymphoma according to the University of California at Davis COAP plus l-asparaginase protocol. One month after initiation of chemotherapy, the cat was admitted to the hospital, quiet, alert and dehydrated with a heart rate of 280 beats/min. Treatment with intravenous fluids (LRS qs 20 mEq/l at 25 ml/h) was initiated. Eighteen hours after presentation no urine had been produced and results from blood drawn at admission showed severe hyperkalemia (10.0 mEq/l; normal range 3.9–5.3 mEq/l); blood measured on a portable analyzer at this time confirmed hyperkalemia (8.6 mEq/l; normal range 2.9–4.2 mEq/l). The initial urine sample was clear in color, with a specific gravity of 1.013, and a negative sediment. Severe azotemia and anuria despite volume expansion suggested acute anuric renal failure possibly due to doxorubicin toxicity. An ECG revealed rapid wide-complex tachycardia and the absence of P waves (Fig 3a). Subsequent to treatments with intravenous fluids (0.9% NaCl with 12 mEq/l sodium bicarbonate at 30 ml/h), and 50% dextrose (Phoenix Laboratories) (1 g/unit of insulin, IV) with regular insulin (0.2 units/kg body weight, IV) the cat started to urinate small amounts. To promote diuresis, furosemide (2 mg/kg body weight, IV) was administered. Sucralfate (Carafate; Axcan Scandipharm) (50 mg/kg body weight, PO, q 8 hours), ondansetron (Zofran; Glaxo Smith Kline) (0.1 mg/kg body weight, IV, q 8 hours) and cimetidine (Tagamet; Glaxo Smith Kline) (10 mg/kg body weight, IV, q 8 hours) were also administered.

Fig 3.

(a) Nine-lead electrocardiogram from case 3 demonstrating wide-complex tachycardia and absent P waves. The lead labeled V1 is rV2 and the lead labeled V3 is V4. Heart rate=300 beats/min. Serum potassium=10.0 mEq/l. Paper speed=25 mm/s. Standardization is 1 cm/1 mV. Cat is in right lateral recumbency. (b) Lead II rhythm strip from case 3 after 15 h of treatment. A P wave is present for each QRS complex. The rhythm is sinus tachycardia. Heart rate=175 beats/min. Serum potassium=7.5 mEq/l. Paper speed=25 mm/s. Standardization is 1 cm/1 mV. Cat is in right lateral recumbency.

Twenty-four hours after admission, the cat had started to urinate larger amounts. The serum potassium level decreased (7.5 mEq/l; normal range 2.9–4.2 mEq/l) and the ECG (Fig 3b) showed sinus tachycardia. Ten days after presentation, the cat was discharged. The cat underwent a thoracotomy 13 weeks after admission; a mediastinal mass was resected and confirmed as a thymoma. Twenty months after presentation for acute renal failure, the cat showed no further clinical signs, although mild azotemia persisted.

Discussion

The cardiotoxic effects of potassium on animals have been known since at least 1839, when Blake found that a concentrated solution of ‘carbonate of potass’, when given intravenously to a dog, ‘arrested the action of the heart in 15 seconds after its injection’ (Blake 1839). This association between severe hyperkalemia and its cardiac effects is of diagnostic value in clinical practice, and indeed ‘the electrocardiogram may be helpful in establishing a suspicion of hyperkalemia while awaiting results for serum potassium concentration’ (Feldman and Nelson 1996, DiBartola and de Morais 2000). Like dogs, hyperkalemic cats typically have heart rates that are in the low or low–normal ranges for a severely ill animal and have electrocardiograms with the traditional hyperkalemic alterations. The concurrent existence of severe hyperkalemia and tachycardia in cats is widely recognized through anecdotal reports, but surprisingly investigations of the ECG characteristics of tachycardia in naturally hyperkalemic cats are few in number (Chamberlain et al 1939, Schaer 1977, Harpster 1987, Macintire 1997, Tilley 1992, Kirby et al 2000, Côté and Ettinger 2005).

The traditional electrocardiographic changes seen with hyperkalemia are due to changes in the action potential and resting membrane potentials. The first recognized cardiac effects of hyperkalemia are slowing of the heart rate, peaking and narrowing of the T wave and shortened QT. In humans, the T waves are typically upright or positive in standard inferior limb leads; however, in cats, the T waves maybe positive or negative. These effects of hyperkalemia on T waves are due to depressions in cell excitability and conduction velocity. As the potassium increases, the ECG demonstrates a widening QRS and flattening P wave due to slow atrial and ventricular conduction. Higher extracellular potassium concentrations decrease the atrial and ventricular muscle's resting membrane potentials, which inactivates sodium channels and decrease both V_max and conduction velocity (Rubart and Zipes 2005). Ultimately, the P wave disappears and the widening QRS complex blends with the T wave in a sine-wave morphology. Aberrant ventricular conduction such as heart blocks, idioventricular complexes and escape beats can also be seen as a result of A–V junctional delay, conduction delays in the His–Purkinje system and delays in ventricular muscle conduction (Schaer 1977). Finally, the hyperkalemia leads to ventricular fibrillation or asystole.

The ECGs of the three cases presented here demonstrate a heart rate>200 beats/min in the presence of severe hyperkalemia. In all three cases there are no discernable P waves in any of the simultaneously recorded leads and the QRS complexes are wide and monomorphic. The differential diagnosis list for lack of P waves includes atrial fibrillation, atrial standstill (whether due to primary atrial myopathy or hyperkalemia), and P waves that are isoelectric in a given lead. Atrial fibrillation produces an irregularly irregular rhythm and therefore is unlikely in these three cases where the rhythm is regular. The multiple lead tracings in these ECGs demonstrate a lack of P waves in all leads. This finding, which has not been investigated in previous reports, rules out P wave isoelectricity as the reason for the lack of observable P waves. Two other points support hyperkalemia as the sole cause of the absent P waves in these cats' ECGs. First, two of the cases return to sinus rhythm with normokalemia. Second, previous studies on atrial muscle have demonstrated its sensitivity to potassium (Chamberlain et al 1939, Ettinger et al 1974a, 1974b).

The wide QRS complexes in all three cases have a morphology that is similar to those seen with ventricular tachycardia or with hyperkalemia. It is difficult to classify or confirm if the origin of the QRS complexes is supraventricular or ventricular using these surface ECG tracings because P waves are absent, making the assessment of atrioventricular association impossible, and wide, bizarre appearing complexes are expected with either ventricular tachycardia or supraventricular tachycardia with concurrent intraventricular conduction delay. However, for the following reasons the authors consider it unlikely that the rhythm is ventricular in origin. First, experimental and clinical hyperkalemia in animals produces wide and bizarre QRS complexes (Schaer 1977, Tilley 1992). Second, electrophysiological studies have demonstrated that in severe hyperkalemia the rhythm is consistently supraventricular, even when AV dissociation is apparent on the ECG (Vassalle et al 1964, Cohen et al 1971). Third, previous studies have demonstrated that atrial muscle is more sensitive to depression of depolarization in the presence of hyperkalemia than either the SA or AV node; therefore, when the P wave disappears, the SA and AV nodes are still functional (Ettinger et al 1974a, 1974b). This effect is likely explained by the fast sodium channel currents responsible for phase zero depolarization in atrial and ventricular myocytes, when compared to calcium-channel dependent cells of the SA and AV nodes. Fourth, the internodal pathways that conduct impulses from the S–A node to the A–V node are more resistant to hyperkalemia than the atrial myocardium (Rubart and Zipes 2005). These pathways can conduct an impulse from the S–A node to the A–V node termed sinoventricular rhythm when the atrial fibers have become impaired due to the lower resting membrane potential seen with severe hyperkalemia. Therefore, the wide, bizarre, rapid complexes seen with hyperkalemia in cats may well be supraventricular impulses with aberrancy, not premature ventricular complexes (Ettinger et al 1974a, 1974b). Without electrophysiological studies on sick cats, it cannot be concluded with certainty that the rhythm is not ventricular tachycardia; however, the ECGs are consistent with previous studies on humans, cats and dogs in which hyperkalemia coexisted with sinus or ectopic supraventricular tachycardia (Winkler et al 1938, Chamberlain et al 1939, Fisch et al 1964, Pick 1966, Cohen et al 1971).

In this group of hyperkalemic cats, several factors may have contributed to the higher heart rates as compared to the traditional experimental models (Cohen et al 1971, Surawicz 1967a, 1967b, Ettinger et al 1974a, 1974b). First, tachycardia has been associated with pain, signs of which have been noted in aortic thromboembolism (Bonagura and Lehmkuhl 1994) and anuria (Cowgill and Elliot 2000). In the experimental models of potassium infusion, pain was unlikely in the anesthetized animal. In numerous human and dog studies, the onset of congestive heart failure has been associated with an increased heart rate (Malliani and Pagani 1983, Fox 1989, Hasenfuss 1998). Such a process may have occurred in case 1, in which the heart rate was normal initially in the absence of overt signs of heart failure, and then was high despite hyperkalemia when dyspnea was noted together with cardiogenic pulmonary edema. Another important factor which could have increased the heart rate in any of these hospitalized cats is environment. The hospital setting can have substantial hemodynamic effects on many cats, as shown in a study of normal cats subjected to simulated veterinary visits (Belew et al 1999). In this study, the cats subjected to simulated veterinary visits had both elevated blood pressures and heart rates compared to the same parameters at rest. Finally, heightened sympathetic activity could stimulate the SA node which could increase the heart rate in a sinoventricular rhythm.

Other factors that may have contributed to the tachycardia in this report are alterations in other serum electrolyte and mineral concentrations. In previous publications, hypocalcemia and hypomagnesemia were associated with tachycardia (Surawicz 1967a, 1967b, VanderArk et al 1973, Fisch 1973, Havestadt et al 1985, Atkins 1999). Hypocalcemia in most clinical settings does not cause tachycardia alone. However, it can lead to a decrease in the duration of the QRS and an increase in conduction velocity. Similarly, hypomagnesemia alone has no appreciable effect on the ECG; however, it can exaggerate the ECG effects of hypocalcemia (Surawicz 1967a, 1967b, Fisch 1973, Dyckner and Wester 1981). Cases 1 and 3 suffered from urinary obstruction which is associated with ionized hypocalcemia (Drobatz and Hughes 1997). Serum ionized calcium and magnesium were not measured in these cats. Hyponatremia and hypernatremia can also alter the ECG through changes in the resting membrane potential; however, arrhythmias are uncommon (Gibson et al 1974). In cases of hyperkalemia, hyponatremia may aggravate the already depressed conduction, while hypernatremia can correct the depressed conduction caused by hyperkalemia (VanderArk et al 1973). In each of the cases presented here, the serum sodium levels were normal.

In the cases presented here, there were no common treatments for hyperkalemia. An essential step in the treatment of hyperkalemia is to identify and begin treatment for the underlying cause. Once the cause has been identified, there are multiple treatment options. The first option is to enhance the urinary excretion of potassium with the administration of intravenous fluids. The second option is to antagonize the effects of the potassium on cell membranes. Calcium gluconate can be used to restore normal membrane excitability. It does so by decreasing the threshold potential, which normalizes the difference between the resting and threshold potential. The third option is to shift the potassium from the extracellular fluid, where it could reach the heart, to the intracellular fluid. This can be done with two methods. The first is to administer 50% dextrose. Glucose increases endogenous insulin and moves potassium into the cell. Regular insulin can be combined with 50% dextrose for a greater reduction in extracellular potassium; however, hypoglycemia can be a complication. The second approach is to administer sodium bicarbonate. Sodium bicarbonate causes the movement of potassium ions into cells as hydrogen ions leave the cell to titrate the bicarbonate. Finally, the potassium can be removed with dialysis.

Despite the cases described in this report, it is the authors' impression that tachycardia occurs in only a minority of cats with severe hyperkalemia. The importance of considering hyperkalemia in the differential diagnosis of a wide-complex tachycardia is evident from the present cases and previous reports. First, the presence of a rapid, regular tachycardia should not rule out the diagnosis of severe hyperkalemia in the cat. Second, the appearance in any feline ECG of a wide-complex tachycardia without P waves, as seen in Figs 1, 2a, and 3a, should prompt the consideration of severe hyperkalemia. Specifically, a serum potassium level should be measured on cats whose history and physical examination are consistent with disorders causing hyperkalemia and whose ECG shows these features. A third, though unverified point in these cases is that instituting ventricular antiarrhythmic treatment based on these ECG abnormalities may be misguided. If sinoventricular rhythm with aberrancy is responsible for these ECG findings, then administration of ventricular antiarrhythmic drugs such as lidocaine is likely to be ineffective and potentially hazardous, and may delay more important measures aimed at reducing critically high serum potassium levels.

These cases support previous anecdotal and case reports indicating that in cats, tachycardia does not exclude the diagnosis of hyperkalemia. These cases also demonstrate that the differential diagnosis for wide-complex tachycardia on the ECGs of cats includes hyperkalemia. These important facts have clear clinical implications in the triage and initial management of critically ill cats.