65 Trial

Published

Lamontagne F, et al. "Effect of Reduced Exposure to Vasopressors on 90-Day Mortality in Older Critically Ill Patients With Vasodilatory Hypotension; A Randomized Clinical Trial". JAMA. 2020. 323(10):938-949.
PubMed • Full text

Clinical Question

In ICU patients aged ≥65 years with vasodilatory hypotension, what is the effect on 90 day mortality by reducing the vasopressor exposure through permissive hypotension of target MAP of 60-65 mmHg?

Bottom Line

Among ICU patients aged ≥65 years with vasodilatory hypotension, targeting a MAP of 60-65 (i.e., permissive hypotension) did not change mortality at 90 days but lowered exposure to vasopressor medications when compared to usual care.

Major Points

The single-center Rivers Trial (2001) found a mortality benefit from an Early Goal Directed Therapy (EGDT) strategy including several interventions, including MAP >65 mm Hg with vasopressors as needed. There is a dearth of evidence to support this specific MAP target, and subsequent multicenter sepsis studies comparing the protocol from the Rivers Trial versus other care patterns were largely negative (e.g., ProCESS, 2014; ARISE, 2014; ProMISe, 2015). These subsequent studies assessed the entire EGDT protocol versus other standards. Whether a higher MAP target alone provided benefit in vasodilatory shock among adults in the ICU was unclear. The use of vasopressor medications to achieve this MAP might introduce unneeded harm to the patients with shock since they involve shunting of the patients blood away from some areas in the body and risk of arrhythmias.

Published in 2020, the 65 Trial was a open label, multicenter RCT conducted in 65 NHS ICUs in the UK that randomized 2,583 adults aged ≥65 years with vasodilatory hypotension despite fluid resuscitation currently receiving vasopressors to 1) Permissive hypotension targeting a MAP of 60-65 mmHg (n=1,283), or usual care (n=1300). The 90 day all-cause mortality 41% (permissive hypotension) and 44% (usual care). The unadjusted comparison between these groups did not reach significance (OR 0.89; 95% CI 0.76-1.04), but an apparently pre-specified analysis that adjusted for multiple covariates did reach significance (OR 0.82; 95% CI 0.68-0.98). There was no difference in serious adverse events. Interpretation of these statistical endpoints is complicated as the authors seem to have prespecified no adjustments in their primary analysis. The use of covariate adjustment in RCTs is well-established, though interpretation of clinical trial results should follow the prespecified methods. The authors tactfully interpreted their findings and noted that vasopressor use in this population is unlikely to cause harm, and might confer some benefit.

This trial was limited by its open-label design and above-target MAPs in the permissive hypotension group (66.7 mm Hg vs. 72.6 mm Hg). Ultimately, it provides initial evidence that vasopressor use to maintain a MAP ≥65 mm Hg might be reasonable among some patients with vasodilatory hypotension, but should not necessarily be the default strategy.

Guidelines

As of August 2021, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, pragmatic, open-label, randomized, controlled trial
• N=2463
  • Permissive hypotension (MAP 60-65) (n=1221)
  • Usual Care (n=1242)
• Setting: 65 Intensive Care Units in the UK
• Enrollment: July 2017 to March 2019
• Follow-up complete: August 2019
• Analysis: Intention-to-treat
• Primary Outcome: All-cause Mortality at 90 days

Population

Inclusion Criteria

• Aged ≥65 years
• Vasodilatory hypotension as assessed by treating clinician
• Infusion of vasopressors within prior 6 hours; ≥1 hour completed
  • If noradrenaline, then a minimum dose of 0.1 µg/kg/min
• Adequate fluid resuscitation is at least underway
• Vasopressors therapy likely to require ≥6 additional hours

Exclusion Criteria

• Vasopressors being used solely as therapy for ≥1 of the following:
  • Bleeding
  • Acute LV or RV failure
  • Post-cardiopulmonary bypass vasoplegia
• Ongoing treatment for brain injury or spinal cord injury
• Imminent death
• Previous enrollment

Baseline Characteristics

From the permissive hypotension group

• Demographics: 43% female, median age 75 years
• Comorbidities: 46% hypertension, 15% atherosclerotic disease, 11% heart failure, 1.3% chronic renal replacement therapy at admission
• Physiologic parameters: Mean APACHE II score 20.9, mean ICNARC physiology score 23.9, median ICNARC predict risk of death 0.33
• Sepsis status: 22% no sepsis, 30% sepsis, 48% septic shock
• Frailty prior to hospital admission: 66% no assistance, 34% minor/major assistance, 0.7% total assistance with all daily activities

Interventions

Full methods previously published Choice of vasopressor and other therapies was at the desecration of treating clinician.

• Permissive hypotension, target MAP 60-65 mm Hg
  • MAP target reinforced through trial-specific prompts/alarms on intravenous infusion pumps and medical notes
• Usual care

Outcomes

Comparisons are permissive hypotension vs. usual care.

Primary Outcomes

90-day mortality

41% vs. 43.8% (absolute risk difference −2.85; 95% CI, −6.75 to 1.05; P = 0.15)

OR 0.89; 95% CI 0.76-1.04

Secondary analysis, including adjustments for multiple covariates listed at the bottom of Table 3 on page 944

OR 0.82; 95% CI 0.68-0.98

Secondary Outcomes

ICU Discharge mortality

30% vs. 31% (P = 0.66)

Acute hospital mortality

39% vs. 41.5% (P = 0.27)

Cognitive decline (IQCODE Score) at 1 year

Lower means much better, higher means much worse

2.93 vs. 2.80 (difference 0.13; 95% CI 0 to 0.25; P = 0.05)

Median duration to vasopressors

33 hrs vs. 38 hrs

Median norepinephrine equivalents exposure

17.7mg vs. 26.4mg (Difference -8.7; 95% CI -12.8 to -4.6)

Mean MAP while on vasopressors

66.7 mmHg vs. 72.6 mmHg (Difference -5.9; 95% CI -6.4 to -5.5)

Subgroup Analysis

Subgroup analyses were compared using an adjusted logistic regression model, not unadjusted.

Chronic hypertension

No: 43.3% vs. 43.4%

Yes: 38.2% vs. 44.3%

P-value for interaction: 0.047

Sepsis-3

No sepsis: 47.1% vs. 42.5%

Sepsis: 30.8% vs. 37.5%

Septic shock: 44.5% vs. 48.6%

P-value for interaction: 0.06

There were no other significant or nearly significant interactions, by other subgroups, including by age, heart failure status, atherosclerotic disease status, predicted risk of death, or specific infusion agent at randomization.

Adverse Events

Any serious adverse event

6.2% vs. 5.8% (P=NS)

Renal failure: 3.2% vs 2.5%

Supraventricular cardiac arrhythmia: 0.9% v. 1%

Criticisms

• Higher proportion of patients in permissive hypotension group (34% vs. 31%) required dependence assistance prior to enrollment, suggesting a more sick or frail population in the comparator
• Used a "clinical nudge" for intervention on pumps and notes
• Open label design may lead to bias from the clinical team
• Mortality in usual care group was higher than expected

Funding

• NIHR HTA Programme project 15/80/39
• Intensive Care National Audit & Research Centre (ICNARC)
• NIHR Research Professorship award and the NIHR Comprehensive Biomedical Research Centre, grant RP-2015-06-018
• Fonds de recherche du Québec–Santé (FRQS) career award
• UK Critical Care Research Group
• NIHR Clinical Research Networks