Abstract

Amiodarone is a frequently used agent in the Intensive Care Unit. Serious and well-known side-effects include liver toxicity, neurologic toxicity and pulmonary toxicity. We describe a case of acute and severe lower back pain after an intravenous bolus of amiodarone for atrial fibrillation in a critically ill patient, which did not relapse after initiation of a slower continuous infusion. Awareness of this rare side-effect may prevent unnecessary diagnostic tests to assess the cause of the pain. Furthermore, physicians should be aware of the possibility of continuing amiodarone infusion at a lower rate if a patient experiences these symptoms.

Introduction

Atrial fibrillation is a commonly encountered problem in patients in the Intensive Care Unit (ICU).[1] In critically ill patients, the European Society of Cardiology Guidelines for the diagnosis and management of atrial fibrillation advise the use of amiodarone, with or without electrical cardioversion, for rate control with the possibility of cardioversion to sinus rhythm.[2]

Amiodarone is a class III antiarrhythmic, the primary mechanism of action is blocking potassium rectifier currents responsible for the repolarization of the heart during phase 3 of the cardiac action potential. Another effect of amiodarone is blocking of beta-adrenergic receptors (e.g., beta-1), calcium channels and sodium channels. It is frequently used in the ICU to treat and prevent supraventricular tachy-arrhythmias or in life-threatening ventricular arrhythmias.

The most common side-effects are corneal microdeposits, photophobia, optic neuropathy, visual halo’s and brady- and tachyarrhythmias. In the ICU, one may encounter more serious side-effects of amiodarone, such as hypo- or hyperthyroidism, liver toxicity, neurologic toxicity and pulmonary toxicity. Pulmonary toxicity usually occurs in the first year of the treatment and most commonly resembles interstitial lung disease, but can also present as acute respiratory distress syndrome, pleural effusion, organizing pneumonia or alveolar haemorrhage.

Phlebitis can be avoided if the patient has a central venous catheter, but is common in peripheral administration. The administration of amiodarone can also lead to cardiac toxicity, but this is related to its therapeutic mechanism.

Drug interactions can also occur because of the inhibiting effect amiodarone has on the cytochrome P450. Another important aspect of amiodarone is the long half-life of several weeks.[3]

We describe a case of acute and severe lower back pain after an intravenous bolus of amiodarone for atrial fibrillation in a critically ill patient, which did not relapse after initiation of a slower continuous infusion.

Case history

A 78-year old woman, with a medical history of paroxysmal atrial fibrillation, hypertension, chronic obstructive pulmonary disease and obesity, was admitted to the ICU due to a saddle pulmonary embolism with right ventricular strain, most likely secondary to a COVID-19 infection. She was started on rivaroxaban and discharged to the pulmonary care unit on day three, from where she was discharged home after an additional five days of admission. Two days later, she returned to the Emergency Department with progressive dyspnoea and hypoxemia. Computed tomography of the chest showed a persistent saddle pulmonary embolism with progressive dilatation of the right ventricle. She was readmitted to the ICU and started on high flow nasal oxygen therapy. Rivaroxaban treatment was discontinued and she was treated with unfractionated heparin. She developed atrial fibrillation with fast ventricular rate, for which a bolus of intravenous amiodarone (150 mg in 20 minutes) was started. Approximately five minutes after initiation of infusion, she complained of acute and severe lower back pain without radiation to the legs. Simultaneously she developed hypotension, with a mean arterial pressure of 55 mmHg. At the time of onset of these symptoms, no other drugs were administered. The cardiac rhythm did not change during the amiodarone infusion. Her abdomen was non-tender and the arterial femoral pulsations were intact. There were no other signs or symptoms, specifically no flushing sensation, anxiety or urticarial rash. She received five milligrams of intravenous morphine and the amiodarone infusion was stopped. The pain completely resolved within minutes. The attending physician was not aware of acute lower back pain as a possible side effect of amiodarone and due to persistent arterial fibrillation, a continuous amiodarone infusion of 120 mg/hour was started, upon which the patient developed no symptoms. The heart rhythm converted to sinus rhythm and amiodarone dose was tapered and eventually discontinued ten hours after initiation. After a total hospital stay of 18 days she was discharged home with high dose low molecular weight heparin and home oxygen therapy due to persistent hypoxemia. Unfortunately, she was readmitted within a week with hypovolemic shock due to bilateral iliopsoas muscle haemorrhage. She developed progressive shock with multi-organ failure despite massive transfusion of blood products and vasoactive medication. In consultation with her family, further medical treatment was deemed futile and not in line with what the patient would have wanted and therefore withheld. She passed away the day after admittance.

Discussion

To the best of our knowledge, only six case reports or case series, describing a total of  nine patients with acute lower back or abdominal pain after infusion of intravenous amiodarone, have been published to date.[4–9] All these reports describe a similar pattern of pain as in our patient, with onset of acute and severe pain within minutes of administration of amiodarone, which completely resolved within 15 minutes after cessation of drug administration. The cases differ in accompanying complaints, with some cases describing a flushing sensation[5], anxiety[7], hypertension[7], hypotension[8], an urticarial rash[4] or no other symptoms[6]. Interestingly, the doses leading up to the complaint greatly varied from patient to patient, with one report describing pain with as little as a total of 5 milligrams infused. Some reports describe symptoms after intravenous but not oral administration[5,7] and one case describes symptoms after the third but not first or second administration of amiodarone.[7] Our report is the first to describe pain after a bolus but not slower continuous infusion of amiodarone. Our patient developed symptoms after infusion of 30 to 40 mg of amiodarone with a rate of 7.5 mg/minute, while infusion at a lower rate of 2 mg/minute did not result in a relapse of symptoms.

Authors of previous case reports have speculated on the cause of the reported pain after amiodarone administration. Some have suggested hypotension with transient mesenteric ischemia as a cause.[8] Polysorbate-80, one of the solvents of intravenous amiodarone, is deemed to induce hypotension due to release of histamine in canines.[10] Since two case reports describe acute pain following intravenous but not oral administration of amiodarone, this solvent has been identified as a possible causative agent.[5,7] Our patient also developed simultaneous hypotension, but this is unlikely to be the sole explanation, since other reports did not describe changes in blood pressure[4–6] or described hypertension.[7] Furthermore, infusion of amiodarone but not its solvents polysorbate-80 and benzyl alcohol caused vasodilation in humans.[11] At this point in time, we have no explanation for the relation between amiodarone and back pain.

This case adds to the current knowledge of amiodarone side effects. Since amiodarone is a frequently used drug in the ICU and atrial fibrillation is often observed, physicians should be aware of the possibility of continuing amiodarone infusion at a lower rate if a patient experiences symptoms of acute and severe lower back pain. Additionally, awareness of this rare side effect may prevent unnecessary diagnostic tests to assess the cause of these symptoms.

Disclosures
All authors declare no conflict of interest. No funding or financial support was received.

Informed consent was obtained from the patient for the publication of this case report.