Utility of vasopressin and steroids in cardiac arrest

In Review

Mentzelopoulous SD, et al. "Vasopressin, steroids, and epinephrine and neurologically favorable survival after in-hospital cardiac arrest: A randomized clinical trial". JAMA. 2013. 310(3):270-279.
PubMed • Full text • PDF

Clinical Question

In adult patients that suffer an in-hospital cardiac arrest, does the addition of vasopressin and corticosteroids to during CPR to the standard epinephrine improve clinical outcomes of Return of Spontaneous Circulation (ROSC) for at least 20 minutes and survival to discharge with favourable neurological function.

Bottom Line

Despite overall outcome data being poor, the addition of vasopressin and corticosteroids to standard epinephrine during cardiopulmonary resuscitation (CPR) appear to improve rates of survival and improved neurological disability, more study is required for long-term outcomes.

Major Points

In-hospital cardiac arrest leads to significant mortality, survival to hospital discharge is ~20%. Amongst those survivors, rates of severe neurological outcomes can range to as high as 25-50%. A previous small study (single centre, n = 100) by the same authors demonstrated improved survival in refractory in-hospital arrest. From this and previous studies the authors proposed to test if the addition of vasopressin in addition to standard epinephrine during the arrest and corticosteroids during the first round of CPR and in response to shock during the post-arrest period would improve mortality and lead to favourable neurological outcome.

Set in three Greek hospitals, 268 patients were randomized to receive Epinephrine + Vasopressin + Methylprednisolone with the first round of CPR followed by Epinephrine + Vasopressin with each subsequent round of CPR compared to standard Epinephrine + placebos. In the intervention group, 18% more patients had Return of Spontaneous Circulation lasting longer than 20 minutes (NNT 5) and 8.8% more had survival to discharge with favourable neurological outcomes (NNT 11). In general, the intervention group needed less CPR and had shorter duration of ALS. However, out to 1 year there was no difference in groups for favourable neurological outcomes.

This is a complicated trial with multiple interventions in both arms. Bias may have existed as inhaled steroids was present and not controlled for as well as pre-arrest vasopressor support. The low NNT is encouraging in the short term (20 minutes up to 60 days) but positive neurological outcomes a year out is not different.

Guidelines

As of May 2018, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, double-blind, parallel group, randomized control trial
• N= 268
  • Vasopressin-steroids-epinephrine (VSE) (n=130)
  • Epinephrine and placebos (n=138)
• Setting: 3 Tertiary Care Centers in Greece
• Enrollment: September 1, 2008, to October 1, 2010
• Mean follow-up: 60 days
• Analysis: Per-Protocol
• Primary outcomes: ROSC for ≥ 20 minutes, 60 day neurologically favorable survival (Cerebral Performance Category (CPC) score 1-2)

Population

Inclusion Criteria

• In-hospital vasopressor-requiring cardiac arrest
  • Defined by European Resuscitation Council guidelines for resuscitation 2005.

Exclusion Criteria

• Age <18 years old
• Terminal illness with life expectancy < 6 weeks
• Do-Not-Resuscitate status
• Arrest due to exanguination
• Arrest before hospitalization
• Treatment with IV steroids before arrest

Baseline Characteristics

Control group displayed

• Demographics: mean Age 63 years, Male 63.8%, mean BMI 25.4
• Comorbidities: Hypertension 55.8%, Coronary artery disease 31.9%, Diabetes 21.7%, Peripheral vascular disease 17.4%, Cardiac arrhythmia 19.6%, Cardiac arrhythmia 19.6%, Valvular heart disease 13%, Cardiac conduction disturbance 4.3%, "non-cardiovascular comorbidities" 55.8%
• Primary Hospitalization: Cardiac 26%, Respiratory 28%, Digestive 25%, Neurological 8%, Trauma 9%, Malignancy 9%, Acute renal 6.5%
• median hospital stay prior to arrest 2
• Cause of arrest: Hypotension 37%, Respiratory depression/failure 38%, Myocardial ischemia 17%, Metabolic abnormality 15%,Arrhythmia 8%
• Initial Rhythm: Ventricular fibrillation/tachycardia 16.7%, Asystole 70.3%, Pulseless electrical activity 13.0%

Interventions

• Epinephrine 1mg + Placebo per CPR cycle
• Epinephrine 1mg + Vasopessin 20 units per CPR cycle (to Max 100 units) + Methylprednisolone 40mg during first CPR cycle
• if develop shock receive Hydrocortisone 100mg IV every 8 hours

Outcomes

Comparisons are Control vs. Vasopressin-Steroid-Epinephrine group.

Primary Outcomes

ROSC for ≥ 20 minutes

65.9% vs. 83.9% (Difference 18% [Odds Ratio 2.98, 95% CI 1.39-6.40] P = 0.003) NNT 5

Survival to discharge with favourable CPC score

5.1% vs. 13.9% (Difference 8.8% [Odds Ratio 3.28, 95% CI 1.17-9.20] P = 0.02) NNT 11

Secondary Outcomes

CPR Cycles, median

5 vs. 4 (P = 0.001)

Advanced Life Support duration, median, min

19 vs. 13 (P = 0.01)

Epinephrine Dose, mean, mg

5 vs. 4 (P = 0.002)

Bicarbonate Dose, mean, mg

90 vs. 68 (P = 0.38)

ROSC for ≥ 20 minutes after first vasopressor dose

13% vs. 19.2% (P = 0.57)

ROSC for ≥ 20 minutes after second vasopressor dose

20.3% vs. 33.8% (P = 0.04)

Subgroup Analysis

Neurological favourable survival at 1 year

3.6% vs. 8.5% (Difference 4.9% [Odds Ratio 2.46] P = 0.13)

if Survived ≥ 4 hours, control (n=73) compared to VSE (n=76), then Survival to discharge with favourable CPC score

8.2% vs. 21.1% (Difference 12.9% [Odds Ratio 3.74, 95% CI 1.20-11.62] P = 0.02) NNT 7

Adverse Events

No adverse event, as listed in the supplement, showed a difference between groups.

Criticisms

• Trial complicated, using multiple interventions in both arms
• Discharge diagnoses did not include any neurological disorder in 44/96 (45.8%)
• Vasopressin and steroids was given in pre-filled syringes and epinephrine supplied in ampules, being the opposite in availability in many centres this may affect the real-world application of these finding and time to administration
• Non-cardiac comorbidities were not clarified
• Pre-existing inhaled corticosteroid 11.6% control vs. 7.7% VSE may have affected outcome
• Study design may have created bias of vasopressin affect with addition of steroids
• Pre-arrest vasopressor support was not determined
• long-term follow-up was not performed

Funding

• Greek Society of Intensive Care Medicine
• the Project “Synergasia” (ie, Cooperation) of the Greek Ministry of Education (09ΣYN-12-1075)