Controversy continues after the publication of two multicenter, randomized, controlled trials. VITAMINS evaluated patients in septic shock and compared standard of care plus hydrocortisone with combined hydrocortisone, vitamin C, and thiamine (HAT) therapy. (JAMA. 2020. doi: 10.1001/jama.2019.22176. [Epub ahead of print].)
CITRIS-ALI evaluated vitamin C for sepsis-induced acute respiratory distress syndrome (ARDS). (JAMA. 2019;322:1261; http://bit.ly/2HdTX6X.) A concise review is needed to appreciate their inability to report the benefit of HAT therapy.
Early resuscitative therapy in septic shock is distinct from other interventions—delayed correction of deteriorating physiology compromises outcomes. A recent study reported profound benefits of HAT if given within six hours of presenting to the ED, with no differences if initiated beyond 12 hours. (Crit Care Shock. 2020;23:23; https://bit.ly/2XGqvQE.) In VITAMINS, time from presentation was not measured or reported, only time from meeting the study criteria. A conservative estimate of HAT delay from presentation in VITAMINS is at least 18 hours, and it simply found a similar non-impact of delayed therapy. Resuscitative use was not assessed.
In CITRIS-ALI, the original analysis mandated by the reviewers inexplicably failed to account for the many deaths in the control group during the first 96 hours, ignoring the rather obvious fact that the worst severity of illness is death. (JAMA. 2019;322:1261; http://bit.ly/2HdTX6X.) This clearly led to the lack of difference in the primary outcome and influenced multiple secondary outcomes, given that biomarkers were not measured in dead patients. What could not be obscured was the statistically significant difference in mortality. A recent letter published in JAMA also demanded an analysis accounting for these early deaths. (JAMA. 2020;323:792.) The study authors complied and reported the primary outcome of SOFA score to be statistically significantly decreased at 96 hours. Thus, a positive study?
In the history of critical illness, multicenter, randomized, controlled trials, the holy grail of evidence-based medicine, have never reported a reproducible mortality benefit from any pharmacotherapy. (Crit Care Med. 2019;47:1680.) The many reasons for its failure as a primary outcome in ICU trials led the NIH to mandate it be a secondary outcome in CITRIS-ALI. (Crit Care Med. 2010;38[10 Suppl]:S534.) Our opinion is that the inexplicable rationale behind the primary analysis, along with mortality as a secondary outcome, led to the publication of CITRIS-ALI as a negative study. This unfortunate result occurred in retrospect just prior to a global pandemic of ARDS and may be contributing to its overwhelming morbidly and mortality.
We question the therapeutic nihilism of Rory Spiegel, MD, in his article, “Magic Bullets Rare in Medicine, Especially for Septic Shock.” (EMN. 2020;42:9; https://bit.ly/2UAlTcY.) The view that secondary outcome differences are due to random error while positive outcomes are “statistical technicalities” in VITAMINS was noted immediately after he described improved SOFA scores, a finding that predicts a more elusive outcome to obtain: mortality.
The overly skeptical stance taught within journal clubs during training leads to a propensity to send interventions to the scientific graveyard rather than finding therapies that may benefit patients. It is unclear why strong scientific plausibility, extensive animal data, and increasing clinical data supporting vitamin C are ignored. While magic bullets are rare, this review indicates vitamin C may be one.
Micah T. Long, MD
Pierre Kory, MD
Paul Marik, MD