Symptomatic digoxin toxicity in a patient on haemodialysis

Abstract

We present a man undergoing regular haemodialysis sessions, who presented with non-specific symptoms of nausea, vomiting and light-headedness. He was found to have significantly raised serum digoxin concentrations, as well as a heart rate of 30 beats per minutes. An ECG showed complete heart block. He has a history of non-ischaemic dilated cardiomyopathy with resistant supraventricular and ventricular tachycardias and was on concomitant beta-blockade and digoxin. On questioning, he reported a gradual decline in his residual urine output over the past 6 months. He was reviewed by the cardiology team and required both pharmacological therapy for reversal of digoxin toxicity and temporary pacing in view of significant bradyarrhythmias. The beta-blockade and digoxin were discontinued. He was kept on continuous monitoring at the Cardiac Critical Care Unit. His symptoms resolved spontaneously once digoxin-specific antibody fragments were administered and temporary pacing successfully performed.

Background

For a number of years, use of cardiac glycosides for the management of supraventricular arrhythmias and for optimisation of congestive heart failure have shown encouraging outcomes. Digoxin is a cardiac glycoside which is readily available and has proven effective; however, its narrow therapeutic window, and its predominant excretion through the urinary pathway, has made its use in patients with end-stage kidney disease (ESKD) and on renal replacement therapy more challenging with increased risk of toxicity. The literature repeatedly recommends caution with its use. Chronic digoxin toxicity may be managed with supportive measures; however, if potentially life-threatening or life-threatening toxicity ensue, use of digoxin-specific antibody fragment (Fabs) is recommended.

Case presentation

We present a man aged 70 years old who undergoes thrice weekly haemodialysis at our renal unit. He has a history of non-ischaemic dilated cardiomyopathy with left ventricular function less than 25%. He also has a history of resistant supraventricular and ventricular tachycardia requiring concomitant use of digoxin and beta-blockade. He has been established on dialysis for the past 3 years. One year prior, he had suffered a cardiac arrest with ventricular tachycardia, following a dialysis session. Although he fit the criteria for primary prevention against further ventricular arrhythmias with an implantable cardioverter defibrillator (ICD), in view of his multiple comorbidities and overall prognosis, his caring cardiology team opted for medical optimisation at the time. He was not a candidate for cardiac resynchronisation therapy in view of a narrow QRS complex.

The patient presented to the emergency department reporting a week history of unremitting epigastric pain, nausea, vomiting and dizziness. He had a systolic blood pressure measuring 120 mm Hg and diastolic blood pressure of 85 mm Hg at presentation. A bedside pin-prick random blood glucose was 5 mmol/L.

Investigations

On examination, he was noted to be bradycardic with a variable heart rate ranging between 30 beats per minute (bpm) and 45 bpm. He had no focal neurological signs. An ECG showed slow atrial fibrillation with runs of junctional rhythm and intermittent complete atrioventricular heart block (figure 1).

Figure 1.

ECG showing slow atrial fibrillation and runs of junctional rhythm without any visible P waves suggestive of atrial fibrillation with complete atrioventricular block.

He was taking carvedilol 12.5 mg two times daily, digoxin 62.5 μg daily and was on lifelong anticoagulation. He had been taking digoxin for over 2 years and his last serum digoxin concentration was 1.1 ng/mL (laboratory reference range: 0.9–2 ng/mL). He reported a recent gradual decline in his residual renal function, rendering him almost anuric. He denied starting any recent new medications or over-the-counter remedies. In view of the profound bradycardia, his beta-blockade and his digoxin were stopped and digoxin levels were measured. The serum digoxin concentration (14 hours postdose) was elevated at >6.4 ng/mL (laboratory reference range: 0.9–2 ng/mL). His serum potassium was 3.86 mmol/L (normal range: 3.5–5.1 mmol/L) following a dialysis session that morning.

Treatment

He was transferred to the cardiac unit for continuous cardiac monitoring. He was given 0.5 mg of atropine intravenously. Fabs (40 mg of DigiFab) were administered, followed by the insertion of a transvenous temporary cardiac pacemaker.

Following Fab administration and pacing, his symptoms improved spontaneously. He was kept in the cardiac unit to monitor for arrhythmias and determine when his native rhythm became once again pacemaker independent. His response to therapy was monitored clinically by assessing his rhythm and heart rate, blood pressure and by use of bedside telemetry. Serum digoxin levels were measured after 24 hours; however, as expected, these remained elevated, since the assay used measures both the Fab-bound and unbound digoxin. The temporary pacemaker was removed after 72 hours and carvedilol was reintroduced at a lower dose (3.125 mg two times daily) while keeping the patient on continuous cardiac monitoring to observe for bradyarrhythmias and tachyarrhythmias. His ECG now showed atrial fibrillation at 64 bpm with no significant pauses.

Outcome and follow-up

The patient was discharged from hospital on low-dose carvedilol. His serum digoxin concentration on discharge a week later was 1.2 ng/mL (reference range 0.9–2.0 ng/mL). Digoxin was discontinued permanently. An outpatient 7-day ambulatory cardiac monitor and repeat echocardiogram was planned for consideration of an ICD, depending on the findings of his forementioned investigations and overall prognosis.

Unfortunately, 1 month following hospital admission, the patient passed away after an out-of-hospital cardiac arrest.

Discussion

Digoxin use has declined over the past two decades, partly because of novel medications for atrial fibrillation and congestive heart failure. Its unique ionotropic and chronotropic properties, however, still make it an attractive option for the management of atrial fibrillation and heart failure, and it has repeatedly proven to be a cost-effective medication.

Digoxin clearance varies linearly with the glomerular filtration rate.¹ In general, patients with an estimated glomerular filtration rate of less than 30 mL/min/1.73 m² are at highest risk for digoxin toxicity. A reduction in digoxin dose (both loading and maintenance) in patients with chronic kidney disease is recommended, as well as more rigorous serum digoxin concentration measurements.

In patients on haemodialysis with a similar background as our patient, digoxin should be used with extreme caution. It should be reserved for patients who have failed rhythm control with other agents or are intolerant to alternative antiarrhythmic drugs. Guidelines generally recommend using lower doses or prescribing digoxin on alternate days for patients on haemodialysis, but it is at the discretion of the caring physician or nephrologist to advocate close monitoring and constitute dose changes.

Digoxin has a narrow therapeutic window and therefore requires dose adjustments when there are changes in the volume of distribution, such as ageing or decrease in adipose stores, and changes in absorption and elimination. This makes its administration and drug monitoring more challenging to the physician. Serum digoxin concentrations should be measured 6 to 12 hours after digoxin administration and after 12 to 24 hours in patients with ESKD or dialysis-dependent individuals.

Cardiac glycosides, such as digoxin, work by inhibiting the Na+/K+ ATPase pump on the cardiac myocyte membrane, resulting in an increase in intracellular sodium, which in turn gives rise to increased calcium concentrations via the sodium–calcium exchanger. The increased calcium concentration prolongs the action potential and increases vagal tone, diminishing conduction at the sinoatrial and atrioventricular nodes, thereby decreasing the heart rate and promoting cardiac contractility when binding to troponin C.²

Potassium and digoxin compete for binding at the same ATPase site; therefore, hyperkalaemia decreases the effect of digoxin, whereas hypokalaemia predisposes the patient to digoxin toxicity. Patients receiving regular haemodialysis sessions tend to have fluctuations in their potassium levels and states of hypokalaemia enhance digoxin toxicity.³

The major route of elimination of digoxin is by renal excretion⁴ making its use problematic in patients undergoing dialysis. The extravascular volume of distribution of digoxin decreases in patients with renal insufficiency, also contributing to increased risk of digoxin toxicity.

Dialysis-dependent patients with minimal residual renal function or who are on polypharmacy are at an even higher risk for digoxin toxicity. Potential drug–drug interactions may be more pronounced in patients with diminished creatinine clearance. In a retrospective study by Chan et al analysing the association between digoxin prescription in patients on haemodialysis and survival, the authors found increased mortality in patients on haemodialysis on digoxin, compared with matched control non-digoxin users. Lower predialysis potassium levels and higher serum digoxin concentration also correlated with increased mortality.³

Bearing this in mind, our patient had the following risk factors for digoxin toxicity, namely: chronic kidney disease on renal replacement therapy, fluctuations in serum potassium that are both dietary and dialysis related and a recent decline in his residual renal function. We suspect that he gradually accumulated digoxin as his residual output diminished and once the serum digoxin concentration surpassed the recommended therapeutic range, this made him feel unwell with systemic symptoms, subsequently rendering him toxic with clinical sequelae. In retrospect, although the fact that patient was on the lowest possible daily dose of digoxin, more rigorous monitoring of his serum digoxin concentrations may have alerted us to identify chronic accumulation of the drug earlier in this high-risk patient with reduced urinary output.

This case of digoxin toxicity was particularly challenging. The patient’s diminished ability to renally excrete the drug, the long pauses on his ECG and the significantly raised serum digoxin concentration put him at risk of an asystolic cardiac arrest. These factors prompted the urgent insertion of a temporary pacemaker, since pharmaceutical therapy alone with DigiFab would not have guaranteed an adequate response in a timely manner. The dose of DigiFab administered partially neutralised the toxicity. We did not persist with full neutralisation with Fab, given that the temporary pacemaker was inserted.

The use of Fab in digoxin toxicity in patients with ESKD has been considered as an effective method to treat potentially life-threatening toxicity and, in fact, haemodialysis does not eliminate the Fab either. Fab therapy was found to have the same efficacy in dialysis-dependent patients and in those individuals with normal renal function. In dialysis-dependent patients, however, there is increased risk of rebound digoxin toxicity, thus more prolonged and rigorous observation should be advocated after Fab administration.⁵ Serum digoxin concentration assays generally measure both unbound and bound digoxin molecules making their use unreliable after Fab administration. Immunoassays that measure free digoxin levels after Fab administration are recommended; however, these are not readily available in all centres. The clinician, therefore, must rely on clinical assessment and ECG tracing to monitor initial response.

In a systematic review with recommendations from EXtracorporeal TReatments In Poisoning Workgroup by Mowry et al discussing digoxin poisoning, the authors concluded from their data that digoxin is only slightly dialysable and that extracorporeal treatment is unlikely to improve the outcomes of digoxin toxicity, regardless if Fab have been administered or not.⁶ This applies to both modalities: haemodialysis and peritoneal dialysis. The use of therapeutic plasma exchange has been suggested as a method of more efficient clearance of the Fab-digoxin molecules and reduction of rebound toxicity in patients with ESKD; however, stronger evidence is needed to support the use of this modality.⁷

Learning points.

• Digoxin should be used with caution in chronic kidney disease and patients on renal replacement therapy, especially those with reduced residual urine output.

• Patients undergoing regular haemodialysis sessions have fluctuations in their potassium levels with hypokalaemia enhancing digoxin toxicity.

• Digoxin-specific antibody fragments (Fab) are recommended for the treatment of potentially life-threatening digoxin toxicity in patients on haemodialysis.

• Temporary cardiac pacing, combined with Fab therapy, can be performed in patients with digoxin toxicity at risk of an asystolic cardiac arrest.

• Patients on renal replacement therapy with digoxin toxicity should be monitored closely following Fab therapy, as they are at increased risk of rebound toxicity.