Abstract
Protamine sulfate is commonly used to reverse the action of heparin after catheter ablation procedures. Serious protamine-related adverse effect is rare, but its recognition and appropriate management by electrophysiologists and intensivists is important. Direct ventricular fibrillation (VF) soon after a slow infusion of protamine has not been clearly described.
We examined the records of all patients who suffered apparent adverse events after protamine administration in our electrophysiology lab from 2013 to 2018. We describe a series of three patients, all of whom suffered a precipitous fall in arterial pressure followed by VF within minutes after administration of protamine following ablation for atrial fibrillation. The same supplier of protamine was used in all three cases, but they were from different batches. Serum tryptase levels were measured in all cases, immediately post-cardiac arrest and at 2- and 6-h post-event. Immunoglobulin levels were not measured. Two patients recovered after aggressive supportive therapy; the third died despite similar support.
We have encountered three cases of profound hypotension followed by VF soon after administration of protamine. Although protamine is safe in a large majority of patients, these adverse events have led our centre to exercise greater selectivity and caution in its use.
In the world of electrophysiology, protamine sulfate is generally considered to be a safe medication and is routinely administered at the end of catheter ablation procedures to reverse the effect of heparin. However, this arginine-rich polypeptide has the potential to cause major adverse effects through multiple pathophysiological mechanisms. This case series documents serious cardiovascular compromise culminating in cardiac arrest soon after administration of protamine in three patients.
From a pathophysiological perspective, protamine sulfate serious cardiovascular adverse effects are recognized to be due to immune hypersensitivity reactions or from direct cardiodepressor effects. It is also possible that there is an arrhythmogenic effect based on experimental studies.
Due to the potential for major adverse effects, patient observation and monitoring is advised during and after protamine administration.
Introduction
Protamine sulfate is an alkalinic polycationic protein molecule comprising of two-thirds arginine residues and one-third amino acids.1 Since its discovery in 1868, it has been a widely used medication for the reversal of heparin, commonly after cardiac or vascular surgery or in heparinized patients with active bleeding. In the electrophysiology lab, protamine sulfate is administered at the end of many cardiac catheter ablation procedures, particularly for atrial fibrillation (AF) to minimize haemorrhagic complications and expedite mobilization.2
Adverse reactions to protamine can occur and are well described in medical literature. Rapid administration produces hypotension and flushing; immune-mediated mechanisms are generally believed to account for more severe reactions.3 Despite its biological origin, severe reactions to protamine are rare. With slow infusions, cardiac arrest with ventricular fibrillation (VF) has not been described.
Methods
Following a fatal incident during an ablation for AF in 2018 that was attributed to an anaphylactic reaction to protamine, we searched for similar events among the records of all patients who underwent AF ablation in our centre in the previous 5 years, a period during which protamine was administered after almost all AF ablations in our centre. Ethics approval was not required. Appropriate consent from the patients and next of kin was obtained.
Results
During the period studied, 2080 ablations for AF or other left atrial arrhythmias were performed. In addition to the index case, we identified two cases of non-lethal complications after ablation that had close similarities to the index case. All had undergone ablation for AF with uninterrupted warfarin therapy and with intravenous heparin to a target activated clotting time of 350 s. All had received protamine sulfate (protamine sulfate solution for injection, 50 mg in 5 mL, Wockhardt Ltd. Mumbai, India), administered by infusion over a period of 5–10 min in accordance with departmental protocols. No other fatal complication occurred from an AF ablation during the study period.
Case 1
A 72-year-old man with paroxysmal AF presented to the cardiac catheter lab for elective ablation under general anaesthesia. He was of good physical health apart from the symptomatic paroxysms of AF. There was a past medical history of ischaemic heart disease and sinus node disease with percutaneous coronary intervention and implantation of a dual-chamber pacemaker 2 years earlier. Left ventricular (LV) function was preserved.
The ablation procedure was unremarkable. A baseline transoesophageal echo (TOE) confirmed the normality of cardiac structure. Radiofrequency lesions were delivered in a wide antral circumferential pattern to isolate the pulmonary veins. A left atrial roof line was then created, during which AF terminated to sinus rhythm. At the conclusion of the procedure, protamine 100 mg was given intravenously by slow infusion.
After 3 min, the patient became progressively hypotensive despite repetitive doses of metaraminol. Cardiopulmonary resuscitation (CPR) was undertaken and adrenaline administered. At 10 min after the administration of protamine there was VF which was refractory to electrical cardioversion. After 2 additional minutes of CPR a return of spontaneous circulation was achieved. Despite normal ventilation, the patient remained hypoxic. Arterial blood gas analysis revealed hypoxaemia and mixed acidosis; the patient remained hypotensive requiring noradrenaline and adrenaline support. Hydrocortisone 100 mg was given as well as chlorphenamine maleate 10 mg.
A repeat TOE now showed mildly dilated right ventricle and mildly impaired LV systolic function with no pericardial effusion. Serial TOEs over the next 3 h demonstrated progressive right ventricular dilatation. Coronary angiography showed unobstructed vessels; pulmonary angiography showed unobstructed pulmonary arteries and no evidence of embolism. Chest radiography at 2 and 10 h post-arrest showed widespread pulmonary infiltrates, in keeping with acute respiratory distress syndrome and not cardiogenic shock. The LV systolic function was only mildly impaired and stable haemodynamic parameters were maintained at this stage. Methylprednisolone was administered. Hours after the arrest, computed tomography of the head identified a small localized intracranial bleed but with no evidence of encephalopathy or raised intracranial pressures. This was not felt to be the primary cause for the cardiac arrest. Despite maximal support in the intensive care unit, the patient died 12 h later. Post-mortem did not find any cause for the cardiac arrest. Tryptase levels that were sent at two intervals (immediate and at 6 h) were normal. A blood film showed mild leucocytosis.
Case 2
A 60-year-old man with persistent AF in the context of dilated cardiomyopathy with a cardiac resynchronization therapy implantable cardioverter-defibrillator (CRT-D) in situ underwent ablation under conscious sedation. The patient underwent pulmonary vein isolation (PVI) by cryotherapy completed with additional ablation to superior vena cava and coronary sinus with cavotricuspid isthmus, mitral line, and roof linear ablation. The ablation was uneventful, and at the close of the procedure protamine 50 mg was given. Within 3 min of its administration the arterial pressure plummeted to less than 80 mmHg without any patient distress. Frequent premature ventricular complexes progressed to sustained haemodynamically unstable ventricular tachycardia and finally to VF (Figure 1).
Simultaneous surface electrocardiographic recording and arterial haemodynamic trace of Case 2, documenting the onset of marked ST segment elevation, frequent premature ventricular complexes, followed by suppression in the mean arterial pressure prior to ventricular fibrillation cardiac arrest. This all occurred within minutes of protamine administration.
Immediate defibrillation followed by CPR with administration of boluses of adrenaline resulted in restoration of cardiac output. Emergency echocardiography showed no new abnormality and coronary angiography showed no obstructive lesions. After supportive therapy on intensive care unit, the patient eventually made a full recovery.
Case 3
A 67-year-old male patient presented electively for a persistent AF ablation procedure under general anaesthetic with cryotherapy. The procedure was uneventful with PVI by cryotherapy associated with ablation of multiple lines of ablation. Atrial fibrillation organized to atrial tachycardia which terminated to sinus rhythm. The procedure was concluded, and protamine 100 mg was given slowly.
At 10 min after protamine infusion, the arterial pressure fell rapidly to undetectable levels followed by broadening of the QRS complex and then the onset of VF. Defibrillation and CPR were performed, and adrenaline boluses were given intravenously resulting in restoration of circulation. Emergency echocardiography showed globally impaired contractility. Coronary angiography showed no significant coronary disease.
The patient was admitted to the intensive care unit; he required only a 24-h stay and recovered quickly. Tryptase levels sent at time of the arrest proved normal. The blood counts revealed high white cell counts and neutrophilia. The patient remains under follow-up at 4 years and is well with good exercise capacity and constant sinus rhythm.
Discussion
We report three cases of cardiac arrest in the interventional electrophysiology lab, all involving profound hypotension followed by VF and all occurring within 10 min after the administration of protamine. The protamine was administered in accordance with recommendations by slow infusion of 50–100 mg over 10 min or more. None of the patients had any history of hypersensitivity or fish allergy. None had prior lung disease, insulin-dependent diabetes or severe kidney failure. All patients were appropriately anticoagulated with uninterrupted warfarin and with peri-procedural heparin to a target activated clotting time (Table 1).
Patient and procedural characteristics
| Patient | Age (years) | Pre-procedural left ventricular (LV) ejection fraction (%) | Comorbidities | Body mass index (BMI) | Medications | Known allergies | Anaesthesia | Biochemistry results | Total heparin (IU) administered | Time interval from last heparin dose to protamine (min) | Time interval post-protamine infusion to onset of cardiac arrest (min) | Other pharmacological agents given peri-procedure (apart from GA) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 72 | >55% | Ischaemic heart disease (IHD), permanent pacemaker (PPM) for symptomatic bradycardia | 28.8 |
| None | General anaesthetic (GA):
|
| 19 000 | 90 | 10 | Intravenous ondansetron 4 mg, paracetamol 1 g, metaraminol 0.5 mg slow bolus |
| 2 | 60 | 30% | Dilated cardiomyopathy (DCM), cardiac resynchronization therapy-defibrillator (CRT-D) in situ | 38.5 |
| None | Conscious sedation
|
| 17 000 | 60 | 3 | Intravenous morphine (5 mg) |
| 3 | 67 | 55% | Chronic kidney disease (CKD) 3 | 28.4 |
| None | GA:
|
| 15 000 | 60 | 10 | Nil |
| Patient | Age (years) | Pre-procedural left ventricular (LV) ejection fraction (%) | Comorbidities | Body mass index (BMI) | Medications | Known allergies | Anaesthesia | Biochemistry results | Total heparin (IU) administered | Time interval from last heparin dose to protamine (min) | Time interval post-protamine infusion to onset of cardiac arrest (min) | Other pharmacological agents given peri-procedure (apart from GA) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 72 | >55% | Ischaemic heart disease (IHD), permanent pacemaker (PPM) for symptomatic bradycardia | 28.8 |
| None | General anaesthetic (GA):
|
| 19 000 | 90 | 10 | Intravenous ondansetron 4 mg, paracetamol 1 g, metaraminol 0.5 mg slow bolus |
| 2 | 60 | 30% | Dilated cardiomyopathy (DCM), cardiac resynchronization therapy-defibrillator (CRT-D) in situ | 38.5 |
| None | Conscious sedation
|
| 17 000 | 60 | 3 | Intravenous morphine (5 mg) |
| 3 | 67 | 55% | Chronic kidney disease (CKD) 3 | 28.4 |
| None | GA:
|
| 15 000 | 60 | 10 | Nil |
Patient and procedural characteristics
| Patient | Age (years) | Pre-procedural left ventricular (LV) ejection fraction (%) | Comorbidities | Body mass index (BMI) | Medications | Known allergies | Anaesthesia | Biochemistry results | Total heparin (IU) administered | Time interval from last heparin dose to protamine (min) | Time interval post-protamine infusion to onset of cardiac arrest (min) | Other pharmacological agents given peri-procedure (apart from GA) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 72 | >55% | Ischaemic heart disease (IHD), permanent pacemaker (PPM) for symptomatic bradycardia | 28.8 |
| None | General anaesthetic (GA):
|
| 19 000 | 90 | 10 | Intravenous ondansetron 4 mg, paracetamol 1 g, metaraminol 0.5 mg slow bolus |
| 2 | 60 | 30% | Dilated cardiomyopathy (DCM), cardiac resynchronization therapy-defibrillator (CRT-D) in situ | 38.5 |
| None | Conscious sedation
|
| 17 000 | 60 | 3 | Intravenous morphine (5 mg) |
| 3 | 67 | 55% | Chronic kidney disease (CKD) 3 | 28.4 |
| None | GA:
|
| 15 000 | 60 | 10 | Nil |
| Patient | Age (years) | Pre-procedural left ventricular (LV) ejection fraction (%) | Comorbidities | Body mass index (BMI) | Medications | Known allergies | Anaesthesia | Biochemistry results | Total heparin (IU) administered | Time interval from last heparin dose to protamine (min) | Time interval post-protamine infusion to onset of cardiac arrest (min) | Other pharmacological agents given peri-procedure (apart from GA) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 72 | >55% | Ischaemic heart disease (IHD), permanent pacemaker (PPM) for symptomatic bradycardia | 28.8 |
| None | General anaesthetic (GA):
|
| 19 000 | 90 | 10 | Intravenous ondansetron 4 mg, paracetamol 1 g, metaraminol 0.5 mg slow bolus |
| 2 | 60 | 30% | Dilated cardiomyopathy (DCM), cardiac resynchronization therapy-defibrillator (CRT-D) in situ | 38.5 |
| None | Conscious sedation
|
| 17 000 | 60 | 3 | Intravenous morphine (5 mg) |
| 3 | 67 | 55% | Chronic kidney disease (CKD) 3 | 28.4 |
| None | GA:
|
| 15 000 | 60 | 10 | Nil |
The ablation procedure in each of the three patients was uneventful, both from a cardiac and anaesthetic perspective. Haemodynamic stability was observed throughout each case up to the time of protamine administration (Table 2). There had been no observed ventricular arrhythmia before the administration of protamine in any case, and the patients who survived the acute event have not had any ventricular arrhythmia since. Blood samples taken before and during the cardiac arrest did not identify any abnormality of electrolyte levels.
Blood pressure recordings pre-procedure, peri-procedure and up until the point of protamine administration
| Patient | Pre-procedural blood pressure recorded in the elective admissions unit (mmHg) | Average blood pressure peri-procedure (mmHg) | Blood pressure at the point of protamine administration (mmHg) |
|---|---|---|---|
| 1 | 120/75 | 115/70 | 120/70 |
| 2 | 121/78 | 145/96 | 158/83 |
| 3 | 105/63 | 110/72 | 117/74 |
| Patient | Pre-procedural blood pressure recorded in the elective admissions unit (mmHg) | Average blood pressure peri-procedure (mmHg) | Blood pressure at the point of protamine administration (mmHg) |
|---|---|---|---|
| 1 | 120/75 | 115/70 | 120/70 |
| 2 | 121/78 | 145/96 | 158/83 |
| 3 | 105/63 | 110/72 | 117/74 |
Blood pressure recordings pre-procedure, peri-procedure and up until the point of protamine administration
| Patient | Pre-procedural blood pressure recorded in the elective admissions unit (mmHg) | Average blood pressure peri-procedure (mmHg) | Blood pressure at the point of protamine administration (mmHg) |
|---|---|---|---|
| 1 | 120/75 | 115/70 | 120/70 |
| 2 | 121/78 | 145/96 | 158/83 |
| 3 | 105/63 | 110/72 | 117/74 |
| Patient | Pre-procedural blood pressure recorded in the elective admissions unit (mmHg) | Average blood pressure peri-procedure (mmHg) | Blood pressure at the point of protamine administration (mmHg) |
|---|---|---|---|
| 1 | 120/75 | 115/70 | 120/70 |
| 2 | 121/78 | 145/96 | 158/83 |
| 3 | 105/63 | 110/72 | 117/74 |
A systematic assessment took place in all three cases to exclude the more likely triggers for cardiac arrest: acute coronary occlusion, pulmonary thromboembolism, and air embolism were excluded. There was no pericardial, pulmonary, or access-site bleeding in any case. A small intracranial bleed was identified in Case 1 at several hours after the arrest, but this had characteristics suggestive of a secondary phenomenon. There was no evidence of intracranial encephalopathy, cerebral oedema, or raised intracranial pressures, so the haemorrhage was not severe enough to have caused the cardiac arrest. Case 1 also had findings suggestive of pulmonary constriction and acute pulmonary hypertension with acute right heart dilatation but there was no evidence of a significant pulmonary thromboembolism. Protamine can cause acute pulmonary vasoconstriction and pulmonary hypertension, but the acute lung injury could also be explained by the prolonged cardiac arrest.
Pharmacology, toxicology, and immunology of protamine sulfate
Animal studies have shown multi-systemic effect of protamine sulfate including cardio-depressor phenomena with severe hypotension and bradycardia.4 A direct suppressive effect on myocardial contractility has been postulated in addition to protamine’s effect on vascular and autonomic system.5 Non-cardiogenic pulmonary oedema has been described in several publications as well as separate experimental observations of marked pulmonary vasoconstriction in canine models.6 Although protamine reverses the action of heparin, excessive doses may increase the risk of bleeding with paradoxical prolongation of activated clotting time.1
It is postulated that protamine can cause direct suppression of myocardial contractility, but this is based on animal studies where protamine was given without heparin. Due to the avidity with which it binds to heparin, the biological effects of protamine in a heparinized patient are those of the protamine–heparin complex, not those of protamine alone. The protamine–heparin complex is considered to be the primary trigger for the rapid pathophysiological development of pulmonary vasoconstriction and pulmonary hypertension but IgG antibodies also act as a trigger. In fact, immunologic mechanisms are thought to underlie protamine’s most serious adverse effects. Type 1 hypersensitivity reactions and delayed reactions have been observed in several clinical cases and in porcine models.1,7 Human physiologic studies on the heparin–protamine complex also confirm these experimental findings.8 Management therefore includes the use of aggressive fluid resuscitation, alpha-agonist vasopressors and intravenous steroids.
Protamine is manufactured from salmon milt. Allergy to shellfish is common but is not correlated with risk of adverse reaction to protamine; molluscs and crustaceans are phylogenetically distant from fish and other vertebrates. True fish allergy is rare: Levy et al.9 recorded just six patients with fish allergy among 4796 cardiac surgical patients, none of whom reacted adversely to protamine. The immune reaction is typically Type 1 hypersensitivity but may also involve immune-complex related pathways activating the complement system.1 Thromboxane A2 is implicated as a component in the immune cascade as part of the prostanoid reaction.
The curious incident of the tryptase in the aftermath
Serum tryptase starts to rise within minutes after an anaphylactic reaction, falling again to a normal level over 12 h. The degree of rise is proportionate to the severity of the reaction. A raised tryptase level strongly supports a diagnosis of anaphylaxis with a positive predictive value of over 89%, but the sensitivity is variable. Some patients have shown normal tryptase levels despite having had an obvious clinical anaphylactic reaction.10,11 A normal tryptase level therefore does not exclude the diagnosis especially when there are strong clinical findings of anaphylaxis. Guidelines still support early measurement of tryptase levels, ideally within 2 h of suspected anaphylactic events, to avoid false negative results. However, a normal tryptase level may also raise the possibility of an alternative pathophysiological mechanism.
Despite taking blood samples within 2 h, we did not detect positive tryptase levels in any of our three patients. At the time of each of these events, we believed that we were dealing with an anaphylactic reaction. Reviewing the three episodes together, the most fascinating shared characteristic is the event that failed to occur: in all cases, the tryptase level failed to rise, suggesting that these may not be anaphylactic reactions. In the Sherlock Holmes short story ‘The Adventure of Silver Blaze’, a killer is implicated by the failure of a dog to bark.12 Having eliminated many possibilities, we remain uncertain what to deduce from the normal tryptase levels.
Other possible triggers for cardiac arrest
Ablation can stimulate unmyelinated vagal C-fibres or ganglionated plexi to produce transient hypotension and long but self-terminating asystolic pauses.13 These, however are immediate effects that subside within seconds of ceasing to deliver energy, so cannot explain an arrest that occurred after all catheters have been removed. There has been a recent case report of electrical storm following PVI in the context of dilated cardiomyopathy,14 but the time course was very different to the cases that we describe with the onset on the day after the ablation. On current evidence, it is not clear if there is a real potential for ablation of ganglionated plexi to cause harm but clinical trials involving this method have not described any risk of electrical storm. Arrhythmogenic effects of anaesthetic agents cannot explain the events in our patients, as the events occurred without general anaesthesia in one case and during recovery from anaesthesia in the others.
The Kounis syndrome is a recently described form of hypersensitivity reaction that leads to coronary spasm and acute myocardial infarction.15 Although it shares some of the features observed in the cases described in this report, our cases lacked any evidence of spasm on their coronary angiograms, so the criteria for the syndrome were not met. The well-known haemodynamic effects of protamine, or of the heparin–protamine complex raise the possibility that the VF encountered was a consequence of myocardial hypo-perfusion arising from systemic arterial hypotension.
A direct arrhythmogenic effect from protamine is plausible but this theory is poorly supported. Animal models have shown an apparent direct suppressive effect on myocardial contractility, but in these experiments, protamine was given without heparin, creating a situation unreflective of clinical use. Iteg et al.16 using a rat model exposed to protamine without heparin showed it to have effects similar to verapamil in reducing contractility, altering the resting membrane potential and the action potential in a dose-dependent manner. It was postulated that protamine can modify cellular calcium shifts via sarcolemmal ion channels, a characteristic that could promote dispersion of refractoriness and hence ventricular arrhythmias (Figure 2).
Flow diagram illustrating the protamine-related pathophysiological mechanisms that may lead to cardiac arrest. IgE, Immunoglobulin E; IgG, Immunoglobulin G; LV, left ventricular.
Safety of protamine
There are some published reports of cardiac arrest following protamine administration in similar clinical settings.17 There are many reports of severe hypotension following protamine infusion in addition to a report of anaphylaxis and acute pulmonary hypertension with right heart failure.18 Overall, the risk of anaphylaxis after exposure to protamine is estimated to be about 1 in 1500 based on two large studies of about 4700 patients.19,20
Despite these negative reports, the majority of patients exposed to protamine experience no adverse effects and various studies report favourably on the safety of protamine, in the context of catheter ablation procedures.20 A review by Sokowlowska et al.21 confirms that there is longstanding and widespread international recognition of the serious adverse reactions associated with protamine and conclude that at present there is no approved alternative to reverse the action of heparin so continued cautious use of protamine is advised. There is ongoing research into alternative agents, including andexanet alfa and PER977.
Pre-medication with anti-histamines in all patients before administration of protamine is a reasonable but unproved strategy. Avoidance of protamine is also an option with prolonged manual pressure and patient immobilization as an alternative. Because of the possible involvement of thromboxane, administration of COX-2 inhibitors such as indomethacin has been suggested as a method of prevention.
Patient safety
Protamine is widely used in cardiac and vascular surgery. More recently it has become common to use it in the cardiac catheter lab after ablation procedures, particularly for AF. The cases reported here were a very small minority of our ablations, so the apparent risk must be weighed against the value of protamine in protecting against major bleeding, and the existence of an extensive literature establishing it as a safe therapy in a large majority of patients.
The patient entering the cardiac catheter lab is exposed to many non-physiological conditions including the ablation process itself, the sedation or anaesthesia with associated ventilation and manipulation of the heart rate and rhythm. Protamine could easily be seen as a small and benign part of the process. The cases serve as a reminder that there are many potential hazards to patient safety, and that the risk does not end at the moment that the equipment is removed from the vascular system.
Management recommendations
Following these serious adverse events following the use of protamine, we have switched from a policy of routine use of protamine after all AF ablations to a selective pattern of use. We have also switched to slower rates of protamine infusion, not exceeding 50 mg in 10 min in accordance with formulary guidelines. Close patient observation during and post-protamine administration is now conducted. Pre-medication with anti-histamines has not been used routinely in our centre but is considered in selected patients, including those with a known history of fish allergy or a history of hypersensitivity reaction to more than 1 substance.
Conclusion
This report documents a series of episodes of severe hypotension followed by VF arrest within 10 min after the administration of protamine after ablation for AF. Although protamine is safe in a large majority of patients, this apparent association has led our centre to exercise greater caution and monitoring of its use in this setting.
Conflict of interest: M.M.G. has received research funding from and has acted as a consultant for Medtronic, Boston Scientific, Biosense Webster and Cook Medical. All remaining authors have declared no conflicts of interest.
References
NICE CG 183.
