Fowler AA, Truwit JD, Hite RD, et al. 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. JAMA. 2019;322(13):1261-1270.
To determine if vitamin C infusion impacts mSOFA scores and biomarkers of inflammation and vascular injury (CRP and thrombomodulin levels).
The thought process behind vitamin C in sepsis is that vitamin C, aka ascorbic acid, is important for endothelial maintenance and catecholamine and cortisol synthesis, all of which are altered during sepsis. Previous studies have found that vitamin C levels fall during sepsis so additional studies started examining if infusing vitamin C during sepsis had an impact.
The debate around vitamin C has been over a decade in the making. CITRIS-ALI was based on studies that showed there may be a mortality benefit with using vitamin C in severe sepsis. However, a double-blind placebo randomized controlled trial was missing. Enter Fowler et al. CITRIS-ALI was designed based on a previous phase 1 study conducted by the same group examining the safety and tolerability of vitamin C, which found significant improvement in mSOFA scores, CRP, and thrombomodulin (markers of inflammation and vascular injury) when vitamin C was given.
CITRIS-ALI is now the first randomized controlled trial looking at the impact of vitamin C in ARDS caused by sepsis.
Patients age > 18 years with sepsis and ARDS for < 24 hours
- ARDS was defined as Pa02/Fi02 < 300 mmgHG, bilateral opacities on CXR, endotracheal intubation, and worsening respiratory failure without evidence of L atrial HTN.
- Sepsis was defined as a suspected or proven infection and meeting 2 of 4 SIRS criteria.
- Multicenter (7 ICUs in U.S. hospitals), double-blinded, randomized, placebo-controlled trial with 167 patients
Patients were randomly assigned to either vitamin C (50 mg/kg in D5W) or placebo (D5W) every 6 hours for 96 hours.
mSOFA score, CRP, and thrombomodulin levels at 0, 48, 96, and 168 hours
Of the 46 secondary outcomes, 3 became relevant:
- All-cause mortality
- Ventilator-free days
- Hospital-free days
There was no significant difference between vitamin C and placebo in mSOFA scores or CRP or thrombomodulin levels. However, the study did find a statistically significant reduction in mortality, hospital free days, and ventilator free days in the vitamin C group.
- Randomized, double-blind, placebo controlled trial
- Well-balanced study arms
- Little to no patient dropout once enrolled
- Improved mortality benefit with vitamin C
On first glance, this study is summarized as a negative study, with no significant difference between vitamin C and D5W with regards to all of the primary outcomes. The study is presented this way in the abstract and is discussed in a similar fashion on many online forums. However, a key topic of discussion with this study is regarding the secondary outcome of mortality benefit.
The mortality rate by day 28 in the placebo group was 46.3% whereas by day 28 with vitamin C the mortality rate was 29.8% (p=0.03). In a disease process with an estimated mortality rate of 40%, this difference is huge. In addition, this comes with no adverse side effects noted for any of the patients receiving vitamin C. To many pure statisticians, this difference, while significant, is lessened because mortality was not the primary outcome the study was designed to address.
Another question that arises with this secondary finding is if vitamin C had such an impact on mortality, why were changes not seen in the primary outcomes, especially the mSOFA scores. With regards to mSOFA, EMCrit has a fantastic summary discussing the impact of survivorship bias. Basically, in this study when a patient expired they were removed from the study, which selectively removes the sickest patients, especially in the control group which had a higher mortality rate. Thus, of the larger number that survived in the vitamin C group makes the intervention group look sicker on average.
With regard to the biomarkers, this study was meant to be a continuation of the prior phase 1 study, which had shown improvements in CRP and throbomodulin biomarkers. The patients enrolled in this study were very sick; nearly 60% were requiring pressor support and about 40% received steroids by the time the study had started. The key difference between these studies (the first that found a statistically significant difference in these markers and this one) may have been the point in the disease course when the vitamin C had started. Some reviewers of this article have postulated that the disease course may have been too advanced to see a significant difference in the primary outcomes as the body's inflammatory response was already in full swing when vitamin C was started.
Despite no difference in the primary outcomes, this study is showing a significant mortality benefit with using vitamin C in ARDS due to sepsis, with no adverse side effects. As one of the first RCTs, this study is adding significantly to the ongoing debate of the impact of vitamin C in sepsis.