Vitamin C Sepsis

Effect of Vitamin C Infusion on Organ Failure and Biomarkers of Inflammation and Vascular Injury in Patients With Sepsis and Severe Acute Respiratory Failure: The CITRIS-ALI Randomized Clinical Trial

Fowler AA. JAMA; 2019;322(13):1261-1270

Clinical Question

  • In patients with sepsis and acute respiratory distress syndrome, can intravenous administration of vitamin C, compared to placebo, reduce organ failure scores and biomarkers of inflammation and vascular injury?

Background

  • Acute Respiratory Distress Syndrome (ARDS) is commonly associated with sepsis, and continues to have a high mortality (up to 45% in LUNG-SAFE)
  • Animal studies have suggested a role for vitamin C for attenuating the inflammatory response and vascular lung injury
  • A preliminary study in humans in 2014 showed a significant reduction in inflammatory markers and Sequential Organ Failure Assessment (SOFA) scores- leading to this phase 2 trial
  • In the interim, while this trial was being conducted, a retrospective observational study was published in CHEST
    • This paper was based on the preliminary human study above, and showed a significant mortality benefit using a retrospective before and after design
    • Significant debate over the role of vitamin C in sepsis has resulted ever since

Design

  • Randomised, double blinded, placebo-controlled, multi centre study
  • 1:1 computer generated randomisation
  • Blinding maintained by pharmacy at each institution- they prepared identical amber-coloured bags of vitamin C and placebo
  • Randomisation within 24 hours
  • Based on a power of 80% and 2 sided alpha of 0.05, multiple computer simulations were run to determine a sample size of 170 (85 per group)

Setting

  • 7 ICUs in the United States
  • Sept 2014 – November 2017

Population

  • Inclusion:
    • Mechanically ventilated
    • PaO2/FiO2 ratio of <300 mmHg
    • Bilateral opacities on chest X ray
    • Recruitment within one week of clinical insult
    • New/worsening respiratory symptoms without evidence of raised left atrial pressure
    • Suspected or proven infection
    • Positive for 2 out of 4 of Systemic inflammatory Response criteria
  • Exclusion:
    • Vitamin C allergy
    • Age <18
    • Non-English speaking or ward of state
    • Unable to gain informed consent
    • >48 hours since achieving ARDS criteria
    • No patient surrogate or physician committed to full support
    • Home ventilation
    • Home oxygen >2L/min
    • Interstitial lung disease
    • Diffuse alveolar haemorrhage
    • Diabetic ketoacidosis
    • Active kidney stone
  • 1267 patients screened, 170 patients randomised, 167 included in primary analysis
  • Baseline characteristics broadly similar between groups. Notable differences:
    • More black people in placebo group (23% vs 15%)
    • More respiratory sepsis in intervention group (82% vs 70%)
    • More abdominal sepsis in placebo group (16% vs 7%)
    • More acute kidney injury in placebo group (31% vs 25%)
    • Worse P/F ratio in intervention group (189 vs 214)

Intervention

  • 50mg/kg vitamin C (dissolved in 5% dextrose) every 6 hours for 96 hours

Control

  • 5% Dextrose only

Management common to both groups

  • Both groups ventilated according to ARDSNet protocol
  • Fluid managed conservatively according to FACTT Lite protocol, and continued until unassisted breathing or study day 7

Outcome

  • Primary outcome: Difference in mean modified SOFA score (Modified by removing bilirubin), CRP and thrombomodulin at 168 hours
  • No significant difference seen in any of the primary outcomes:
    • SOFA Score differences: 3 point reduction (intervention) vs 3.5 point reduction (control)- 95% CI -1.23 to 1.03, P= 0.86)
    • CRP: 54 vs 46.1 micrograms/ml (95% CI -8.23 to 24.1, P= 0.33)
    • Thrombomodulin 14.5 vs 13.8 ng/ml (95% CI -2.8 to 4.2, P= 0.70)
  • Secondary outcome: 46 prespecified secondary outcomes- out of these, 3 were positive:
    • 28 day mortality (29.8% vs 46.3%, 95% CI 2% to 31%, P= 0.03)
    • ICU free days to day 28 (Median 11 vs 0 days, P= 0.03)
    • Hospital free days to day 60 ( Median 22 vs 0 days, P= 0.04)

Authors’ Conclusions

  • In this preliminary study of patients with sepsis and ARDS, a 96 hour infusion of vitamin C compared with placebo did not significantly improve organ dysfunction scores or alter markers of inflammation and vascular injury

Strengths

  • Well randomised and blinded
  • Registered on clinicaltrials.gov
  • Power calculation based on previous data

Weaknesses

  • Significant numbers of patients were excluded from enrolment, resulting in the trial taking 3 years to complete
    • The most common exclusion reason was being enrolled >48hrs post achieving ARDS criteria- this is either a problem with screening, or a reflection of the difficulty in diagnosing ARDS
  • The primary outcome measure, while based on the previous phase 1 trial, is not one of interest to patients or treating clinicians
    • Apparently this was decided upon by the NIHR, and was not the first choice of the research team (this is based on personal communication with AA Fowler)
  • Patients who died before hour 96 were not included in the primary analysis, essentially giving them an mSOFA score of zero
    • If more patients died in the placebo group, their overall mSOFA scores would be lowered as a result
    • This would have nullified any difference in the primary outcome

The Bottom Line

  • This randomised control trial of vitamin C in sepsis-induced ARDS showed no difference in the primary outcome of reduction in SOFA or biomarkers
  • The secondary outcome of mortality is purely hypothesis forming
  • Vitamin C cannot be recommended for use on the basis of this trial
  • I am not currently using Vitamin C routinely in my septic shock or ARDS patients, and will await the results of upcoming large RCTs on its use (VICTAS, VITAMINS)

External Links

Metadata

Summary author: Segun Olusanya
Summary date: 21/10/19
Peer-review editor: Fraser Magee

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.