Vasopressin Administration in Refractory Cardiac Arrest

Methods

In a final effort to resuscitate patients in whom standard American Heart Association Advanced Cardiac Life Support therapies after in-hospital cardiac arrest had failed, arginine vasopressin was administered centrally as an initial 40-U bolus. Each patient had received standard manual cardiopulmonary resuscitation with at least 1 mg of epinephrine and an attempt at direct-current shock before receiving vasopressin through either a femoral or jugular vein. Table 1 lists patient demographic characteristics, and Table 2 provides some details of therapy and outcome after cardiac arrest. One of eight patients (patient 2) had an unwitnessed arrest. Cardiopulmonary resuscitation was initiated less than 1 minute after arrest in the remaining patients; cardiopulmonary resuscitation and advanced cardiac life support were done on all patients for at least 12 minutes (mean ±SD, 21.6 ± 11.8 minutes) before vasopressin was administered. Three patients were discharged from the hospital with good neurologic recovery.

Case Highlights

Patient 4

Four days after having a hemicolectomy, a 78-year-old woman developed pulmonary emboli and ventricular fibrillation. Defibrillation (200 J) led to asystole. The patient received cardiopulmonary resuscitation and epinephrine (1-mg, 3-mg, and 5-mg doses administered 3 minutes apart). After the 5-mg bolus, ventricular fibrillation evolved but was resistant to repeated direct-current shocks and to lidocaine (100 mg). Central administration of vasopressin (40 U) followed by direct-current shock (360 J) resulted in a supraventricular rhythm with a palpable carotid pulse. A systolic blood pressure of approximately 100 mm Hg was maintained with a norepinephrine infusion of 0.15 µg/kg of body weight per minute. After uncomplicated embolectomy, the patient was transferred to the intensive care unit for 3 days and was discharged without neurologic deficit 4 weeks later.

Patient 5

A 71-year-old woman developed ventricular fibrillation while her chest was being scrubbed before implantation of a permanent pacemaker. Closed-chest cardiac massage was initiated within seconds. After three successive direct-current shocks (200 J, 300 J, and 300 J) followed by epinephrine (1 mg) and another direct-current shock (300 J), the patient remained in ventricular fibrillation. Examination of arterial blood gases showed a pH of 7.33 and a PO ₂ of 60 mm Hg before endotracheal intubation. Additional epinephrine (1 mg) and defibrillation efforts were unsuccessful. Vasopressin (40 U) was administered 50 minutes after the arrest, and spontaneous circulation returned immediately after a 300-J direct-current shock. Immediately before vasopressin administration, the patient's arterial blood had a pH of 7.18 and a PO ₂ of 543 mm Hg. The patient was treated with dopamine (10 µg/kg per minute) intravenously. Forty-five minutes later, she again developed hypotension followed by ventricular fibrillation. Cardiopulmonary resuscitation was reinitiated, but direct-current shock (300 J), epinephrine (1 mg), and another direct-current shock (300 J) failed to revive her. Vasopressin (20 U) followed 30 seconds later by direct-current shock (300 J) was unsuccessful; more vasopressin (20 U) was administered 2 minutes after the first 20-U dose. Thirty seconds later, a direct-current shock (300 J) led to an immediate return of spontaneous circulation. The patient's pulmonary capillary wedge pressure at this time was 25 mm Hg, and her pulmonary systolic pressure was 35 mm Hg. Twenty minutes after her second and final resuscitation effort, she became hypotensive and bradycardic and died secondary to pulseless electrical activity.

Patient 6

Immediately after induction with a standard cardiac general anesthetic for placement of an implantable cardioverter-defibrillator, a 45-year-old man developed pulseless electrical activity. Standard closed-chest manual cardiopulmonary resuscitation was started immediately. The patient received fluids (500 mL of normal saline), atropine (1 mg intravenously), and epinephrine (1 mg intravenously). After 10 minutes and another 1-mg epinephrine dose, he developed ventricular fibrillation. Several efforts to defibrillate failed. Twenty minutes after cardiac arrest, the patient received vasopressin (40 U); after a single 360-J transthoracic direct-current shock, spontaneous circulation promptly returned. The patient remained hemodynamically stable for 30 minutes. Despite intravenous fluids, dopamine (10 µg/kg per minute), and placement of an intra-aortic balloon pump, he again developed hypotension, followed by ventricular fibrillation. After an effort to resuscitate the patient with standard cardiopulmonary resuscitation, epinephrine (1 mg), and direct-current shock was unsuccessful, the patient was given vasopressin (40 U) and was successfully resuscitated with direct-current shock. An angiogram showed a large thrombus at the site of an angioplasty done 2 weeks earlier; the vessel was again dilated. Within 30 minutes, the patient developed polymorphous ventricular tachycardia and had another cardiac arrest. Standard manual cardiopulmonary resuscitation, intravenous vasopressin (40 U), and direct-current shock were not effective. The patient received active compression-decompression cardiopulmonary resuscitation and vasopressin (40 U). Systolic arterial pressure increased to more than 100 mm Hg; when active compression-decompression cardiopulmonary resuscitation was stopped, the patient spontaneously converted—without direct-current shock—to sinus tachycardia. One hour later, ventricular fibrillation again developed. Resuscitation efforts were terminated.

Patient 8

A 31-year-old man had several internal injuries after a car accident. He developed ventricular fibrillation on the way to the operating room for emergent repair of a ruptured aorta. Fibrillation persisted despite many direct-current shocks and the administration of epinephrine (2 × 1 mg repeated after 3 minutes). After 4 minutes of closed-chest cardiopulmonary resuscitation, examination of the arterial blood showed a pH of 7.16, a PCO ₂ of 54 mm Hg, a PO ₂ of 49 mm Hg (fraction of inspired oxygen, 1.0), a potassium level of 2.8 mmol/L, and a hemoglobin level of 9.1 g/L. Despite treatment with epinephrine, diastolic arterial pressures remained less than 15 mm Hg. Administration of vasopressin (40 U) increased the diastolic arterial pressure to 30 mm Hg, and a subsequent direct-current shock (360 J) restored a stable heart and blood pressure. After the operation, the patient was transferred to the intensive care unit.

Discussion

These cases show that in patients in cardiac arrest who are receiving closed-chest cardiopulmonary resuscitation and have not responded to the standard doses of epinephrine recommended by the American Heart Association, spontaneous circulation can be restored by intravenous administration (through the femoral or jugular vein) of 40 U of vasopressin. These results are consistent with recent data from animals showing that vasopressin has greater efficacy than epinephrine during cardiopulmonary resuscitation [1, 2]. Although the prognosis was poor in all cases and all conventional measures had failed, spontaneous circulation was restored in all eight patients after vasopressin administration. Three patients survived to hospital discharge with minimal or no neurologic deficit. In addition, when active compression-decompression cardiopulmonary resuscitation was combined with the use of vasopressin, one patient had spontaneous conversion to sinus rhythm without the use of direct-current shock. Although the optimal dose of vasopressin in humans is not known, 40 U was effective in all of our patients. In one patient, a dose of 20 U was not effective.

In the eight patients studied, an initial dose of 1 mg of epinephrine was administered. In four of these eight patients, an escalating dose of epinephrine (from 1 mg to 3 mg to 5 mg) was used but was similarly ineffective. In humans having cardiac arrest, epinephrine therapy is used on the basis of case reports and animal studies [5, 6]. Recent clinical trials comparing low-dose with high-dose epinephrine show that the latter has no significant advantage [7, 8]. A more recent placebo-controlled trial showed that neither high- nor low-dose epinephrine had benefit compared with placebo [9]. In our patients, vasopressin may have been more effective because of several factors. Vasopressin exerts a greater vasoconstrictive effect under conditions of hypoxia and acidosis than does epinephrine, and the effects of vasopressin last longer [1, 2]. Vasopressin causes a greater increase in arterial tone than does epinephrine, an effect that correlates with greater myocardial perfusion [1, 2]. Finally, epinephrine increases myocardial oxygen consumption and lactate production in the arrested heart, and vasopressin does not.

Despite the potential value of vasopressin, the drug may have potentially deleterious effects during cardiopulmonary resuscitation, although none was apparent in this small number of cases. Vasopressin may impair perfusion of the collateral-dependent myocardium and thereby exacerbate regional ischemia [10]. We do not know what adverse effects vasopressin may have in humans after resuscitation in terms of impaired vital organ function and splanchnic circulation. Our present knowledge does not yet warrant the widespread use of vasopressin in patients with refractory cardiac arrest. On the basis these case reports, however, further investigation of vasopressin in patients having cardiac arrest is under way. A full-scale randomized comparison of vasopressin and epinephrine is needed before vasopressin therapy can be recommended for refractory cardiac arrest.

Dr. Lurie: Cardiac Arrhythmia Center, University of Minnesota Medical School, Box 508 UMHC, 420 Delaware Street Southeast, Minneapolis, MN 55455.