← Back
NeuroTrials.ai
Neurology Clinical Trial Database

Prostacyclin in SAH

Effects of Prostacyclin on Cerebral Blood Flow and Vasospasm After Subarachnoid Hemorrhage: Randomized, Pilot Trial

Share Share Copied! Compare

Year of Publication: 2015

Authors: Rune Rasmussen, MD; Jørn Wetterslev, MD, ..., DMSc

Journal: Stroke

Citation: Stroke. 2015;46:37-41

PDF: https://www.ahajournals.org/doi/epdf/10....EAHA.114.007470

Clinical Question

Does continuous infusion of prostacyclin (epoprostenol) affect cerebral blood flow, delayed ischemic neurological deficits, and vasospasm in patients with aneurysmal subarachnoid hemorrhage when administered during the vasospasm phase (days 5-10)?

Bottom Line

Administration of prostacyclin to patients with subarachnoid hemorrhage appears safe and feasible. Global cerebral blood flow after SAH is not markedly affected by prostacyclin administration in the tested dose range (1-2 ng/kg/min). While this pilot trial was not powered to detect differences in clinical outcomes, the observed reduction in point estimates of DIND (21-23% vs 38%) and vasospasm (17-30% vs 36-38%) in prostacyclin groups compared to placebo may represent a treatment effect worth investigating in larger trials. No dose-dependent benefit was observed in the tested range, and future trials should focus on doses around 1 ng/kg/min.

Major Points

  • First randomized clinical trial investigating prostacyclin effects on human brain after SAH
  • Single-center, double-blind, placebo-controlled, parallel-group pilot trial
  • 90 patients randomized 1:1:1 to prostacyclin 1 ng/kg/min, 2 ng/kg/min, or placebo
  • 264 patients screened, 111 met inclusion criteria, 90 enrolled, 89 analyzed (1 withdrew consent)
  • Intervention initiated day 5 after SAH and discontinued day 10 (peak vasospasm period)
  • Primary outcome: change in global cerebral blood flow from baseline measured by CT perfusion
  • No statistically significant difference in global CBF change between groups (P=0.20)
  • Placebo group: mean CBF change -4.65 mL/100g/min (95% CI -8.64 to -0.17)
  • 1 ng/kg/min group: mean CBF change -2.05 mL/100g/min (95% CI -5.58 to 1.47)
  • 2 ng/kg/min group: mean CBF change -0.066 mL/100g/min (95% CI -3.59 to 3.46)
  • DIND incidence highest in placebo (38%), lowest in 1 ng/kg/min group (21%), but not statistically significant (P=0.28)
  • Moderate-severe vasospasm: 17% (1 ng/kg/min) vs 36% (placebo), but not statistically significant (P=0.24)
  • Risk ratio for DIND with 1 ng/kg/min vs placebo: 0.55 (95% CI 0.23-1.28)
  • Risk ratio for vasospasm with 1 ng/kg/min vs placebo: 0.47 (95% CI 0.19-1.19)
  • No significant difference in 3-month Glasgow Outcome Scale between groups (P=0.46)
  • Adjusted analysis showed trend toward worse outcome in 2 ng/kg/min group (P=0.18)
  • Overall 3-month mortality: 3.3% (1 placebo, 2 in 1 ng/kg/min group)
  • No serious adverse reactions including bleeding or hypotension observed during infusion
  • Similar rate of serious adverse events across all groups (41-46%, P=0.92)
  • No significant difference in endovascular intervention rates (17-27%, P=0.68)
  • CT perfusion at baseline (day 3±1) and during intervention (day 8±1)
  • All scans analyzed by 2 independent blinded reviewers with consensus for discordant results
  • No dose-dependent effect observed - 2 ng/kg/min group did not show superior outcomes
  • Trial designed to investigate effects during vasospasm phase, not primarily prevention or treatment

Design

Study Type: Single-center, randomized, double-blind, placebo-controlled, parallel-group, pilot trial

Randomization: 1

Blinding: Double-blind. Computer-generated allocation list accessible only by project nurses not involved in patient care. Nurses prepared blinded trial medication and allocated intervention by telephone contact. Patients, treating physicians, outcome assessors, and data analysts blinded. All analyses performed while preserving blinding. Abstract with conclusions written and approved before unblinding

Enrollment Period: Not specified (trial approved 2011, registered 2011, published 2015)

Follow-up Duration: 3 months

Centers: 1

Countries: Denmark

Sample Size: 90

Analysis: Modified intention-to-treat analysis (89 of 90 randomized patients - 1 withdrew consent before intervention). Sample size calculated for 25 patients per arm based on type 1 error 5%, type 2 error 20% (power 80%), SD 35%, ability to detect 20% difference in global CBF. Primary analysis without adjustment for baseline covariates; adjusted analyses performed for primary and secondary endpoints adjusting for WFNS grade, Fisher grade, and age. Continuous variables compared using ANOVA, binary outcomes using chi-square test. P<0.05 considered significant for primary outcome; P<0.01 definitely significant and P=0.01-0.05 indicative of significance for secondary outcomes. All analyses 2-sided with maximum type 1 error 5%. No interim analysis performed. No stratification for baseline characteristics. Statistical analysis performed using SAS version 9.4

Inclusion Criteria

  • Age 18-85 years
  • Aneurysmal subarachnoid hemorrhage documented by angiogram
  • Aneurysm treated with endovascular coiling or microsurgical clipping
  • World Federation of Neurological Surgeons (WFNS) score 1-4
  • Fisher grade 3 or 4
  • Informed consent from patient or approved proxy

Exclusion Criteria

  • Previous subarachnoid hemorrhage
  • Pregnancy or lactation
  • Renal failure
  • Heart failure
  • Bleeding diathesis
  • Major complication during endovascular procedure or surgery
  • SAH from posterior inferior cerebellar artery (PICA) aneurysm
  • Unable to provide informed consent

Baseline Characteristics

Placebo:

  • Number: 30
  • Mean age (range), years: 54 (27-75)
  • Women: 23 (77%)
  • WFNS grade 1: 14 (47%)
  • WFNS grade 2: 12 (40%)
  • WFNS grade 3: 0 (0%)
  • WFNS grade 4: 4 (13%)
  • Fisher grade 3: 22 (73%)
  • Fisher grade 4: 8 (27%)
  • Aneurysm treated with surgery: 17 (57%)
  • Aneurysm treated with coiling: 13 (43%)

Prostacyclin 1 ng/kg/min:

  • Number: 30
  • Mean age (range), years: 50 (22-71)
  • Women: 24 (80%)
  • WFNS grade 1: 14 (47%)
  • WFNS grade 2: 11 (37%)
  • WFNS grade 3: 0 (0%)
  • WFNS grade 4: 5 (17%)
  • Fisher grade 3: 22 (73%)
  • Fisher grade 4: 8 (27%)
  • Aneurysm treated with surgery: 18 (60%)
  • Aneurysm treated with coiling: 12 (40%)

Prostacyclin 2 ng/kg/min:

  • Number: 30
  • Mean age (range), years: 56 (35-77)
  • Women: 27 (90%)
  • WFNS grade 1: 21 (70%)
  • WFNS grade 2: 5 (17%)
  • WFNS grade 3: 0 (0%)
  • WFNS grade 4: 4 (13%)
  • Fisher grade 3: 23 (77%)
  • Fisher grade 4: 7 (23%)
  • Aneurysm treated with surgery: 15 (50%)
  • Aneurysm treated with coiling: 15 (50%)

Arms

FieldProstacyclin 1 ng/kg/minProstacyclin 2 ng/kg/minControl
InterventionContinuous intravenous infusion of epoprostenol (prostacyclin) 1 ng/kg per minute initiated on day 5 after SAH and discontinued on day 10. All patients received baseline CT perfusion at day 3±1, second CT perfusion at day 8±1 during intervention, CT angiography at day 8±1, daily transcranial Doppler ultrasonography, neurological assessment every 4 hours by nursing staff, daily assessment for DIND by principal investigator, and Glasgow Outcome Scale at 3 months follow-up. Standard SAH care including nimodipine administered per institutional protocolContinuous intravenous infusion of epoprostenol (prostacyclin) 2 ng/kg per minute initiated on day 5 after SAH and discontinued on day 10. All patients received baseline CT perfusion at day 3±1, second CT perfusion at day 8±1 during intervention, CT angiography at day 8±1, daily transcranial Doppler ultrasonography, neurological assessment every 4 hours by nursing staff, daily assessment for DIND by principal investigator, and Glasgow Outcome Scale at 3 months follow-up. Standard SAH care including nimodipine administered per institutional protocolContinuous intravenous infusion of drug solvent (placebo) initiated on day 5 after SAH and discontinued on day 10. All patients received baseline CT perfusion at day 3±1, second CT perfusion at day 8±1 during intervention, CT angiography at day 8±1, daily transcranial Doppler ultrasonography, neurological assessment every 4 hours by nursing staff, daily assessment for DIND by principal investigator, and Glasgow Outcome Scale at 3 months follow-up. Standard SAH care including nimodipine administered per institutional protocol
Duration5 days infusion (days 5-10 after SAH), 3 months total follow-up5 days infusion (days 5-10 after SAH), 3 months total follow-up5 days infusion (days 5-10 after SAH), 3 months total follow-up

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Change in global cerebral blood flow from baseline, calculated as CBF during intervention (day 8±1) minus CBF at baseline (day 3±1). Global CBF estimated from average of 6 predefined cortical regions of interest representative of anterior, middle, and posterior cerebral artery territories in both hemispheres. CBF calculated using deconvolution method on Phillips Brilliance CT scanner covering 4-cm slab at basal ganglia level. Analyzed by 2 independent blinded reviewers with consensus reading for >20% discordancePrimaryMean change -4.65 mL/100g/min (95% CI -8.64 to -0.17)0.20 (unadjusted), 0.21 (adjusted for WFNS grade, Fisher grade, age)
Delayed ischemic neurological deficit (DIND) - occurrence of ≥1 DIND during intervention, defined per Vergouwen criteria | Intervention - 1 ng/kg/min: 21% (6/29), 95% CI 10-39%; Intervention - 2 ng/kg/min: 23% (7/30), 95% CI 12-41%Secondary38% (11/29), 95% CI 23-56%RR 1 ng/kg/min vs placebo: 0.55 (95% CI 0.23-1.28); RR 2 ng/kg/min vs placebo: 0.62 (95% CI 0.28-1.37)0.28 (unadjusted), 0.36 (adjusted)
Moderate to severe angiographic vasospasm on CT angiography at day 8±1, qualified by blinded experienced neuroradiologist | Intervention - 1 ng/kg/min: 17% (5/29), 95% CI 8-35%; Intervention - 2 ng/kg/min: 30% (9/29), 95% CI 16-49%Secondary36% (10/28), 95% CI 22-55%RR 1 ng/kg/min vs placebo: 0.47 (95% CI 0.19-1.19); RR 2 ng/kg/min vs placebo: 0.81 (95% CI 0.38-1.71)0.24 (unadjusted), 0.15 (adjusted)
Change from baseline in regional cerebral blood flow in each of 6 vascular territories (right and left anterior, middle, posterior artery territories)SecondarySee detailed CBF data - no significant differences foundNo statistically significant difference for any territoryNot significant for any territory
Unfavorable clinical outcome at 3 months - Glasgow Outcome Scale 1-4 (dichotomized as GOS 5 vs GOS 1-4) | Intervention - 1 ng/kg/min: 28% (8/29), 95% CI 15-46%; Intervention - 2 ng/kg/min: 40% (12/30), 95% CI 25-58%Secondary27% (8/30), 95% CI 14-45%RR 2 ng/kg/min vs placebo: 1.50 (95% CI 0.72-3.14)0.46 (unadjusted), 0.18 (adjusted - trend toward worse outcome in 2 ng/kg/min group)
Mortality at 3 months | Intervention - 1 ng/kg/min: 2/29 (6.9%); Intervention - 2 ng/kg/min: 0/30 (0%)Secondary1/30 (3.3%)Overall mortality 3.3% (3/89)
Endovascular intervention for vasospasm treatment | Intervention - 1 ng/kg/min: 17% (5/29); Intervention - 2 ng/kg/min: 27% (8/30)Secondary23% (7/30)0.68
Elevated transcranial Doppler values (>50% from baseline) in middle cerebral arterySecondaryNot separately reportedNo statistically significant difference between groups0.64
Serious adverse events (≥1 SAE)Adverse43% (13/30)Intervention - 1 ng/kg/min: 41% (12/29); Intervention - 2 ng/kg/min: 46% (14/30)0.92
Serious adverse reactions (bleeding, hypotension)AdverseNone observedIntervention - 1 ng/kg/min: None observed; Intervention - 2 ng/kg/min: None observedNo serious adverse reactions related to prostacyclin observed in any patient

Subgroup Analysis

Adjusted analyses performed for primary and secondary endpoints controlling for predefined covariates: WFNS grade, Fisher grade, and age. Results similar to unadjusted analyses except adjusted analysis showed trend toward unfavorable outcome in 2 ng/kg/min group (P=0.18)

Criticisms

  • Single-center study limits generalizability
  • Small sample size (90 patients) - pilot trial not powered to detect differences in clinical outcomes
  • Only 89 patients analyzed (1 withdrew consent before intervention)
  • CT perfusion scans from 3 patients uninterpretable due to head movement, reducing primary outcome analysis to 86 patients
  • CT perfusion only covered 4-cm slab at basal ganglia level - could not monitor entire brain for small hypoperfused areas
  • Primary outcome (global CBF) is surrogate outcome with unknown association to long-term clinical outcomes
  • Intervention timing (days 5-10) makes it neither purely preventative nor purely therapeutic trial
  • All patients received intervention regardless of DIND symptoms - potential effect might be more pronounced if only symptomatic patients treated
  • Broad confidence intervals due to small sample size limit interpretation of secondary outcomes
  • No formal correction for multiple comparisons in secondary outcomes
  • CT angiography quality insufficient for interpretation in 3 cases
  • Trial design focused on vasospasm phase rather than early prevention
  • Dose range tested (1-2 ng/kg/min) may not be optimal - higher doses not explored due to safety concerns
  • No dose-dependent effect observed in tested range questions dose selection rationale
  • Trend toward worse outcome in high-dose (2 ng/kg/min) group in adjusted analysis (P=0.18)
  • CBF as outcome measure is novel for DIND trials with uncertain clinical relevance
  • Cannot rule out local perfusion effects that global CBF measurement would miss
  • Follow-up limited to 3 months - longer-term outcomes unknown
  • Enrollment criteria (WFNS 1-4, Fisher 3-4) may exclude patients most likely to benefit
  • No assessment of whether intervention affects only symptomatic vasospasm or prevention
  • Timing of baseline CT perfusion (day 3±1) may miss early perfusion changes

Funding

Copenhagen University Hospital research fund, a nonprofit organization with no influence on the trial protocol, conduct, or analysis

Based on: Prostacyclin in SAH (Stroke, 2015)

Authors: Rune Rasmussen, MD; Jørn Wetterslev, MD, ..., DMSc

Citation: Stroke. 2015;46:37-41

Content summarized and formatted by NeuroTrials.ai.