The use of IV vitamin C as a therapeutic agent in sepsis has caught the hearts and minds of emergency medicine and critical care clinicians. Since the publication of a small single center before and after study by Marik et al in 2017 (1), we have spent countless hours screaming into the void that is social media, debating whether the current evidence supports its widespread use. The publication of CITRIS-ALI last year (2), which found a statistically significant increase in survival associated with the use of IV vitamin C, only added to the existing controversy. But the before and after trial design of the original study, and secondary nature of the findings in CITRIS-ALI has given the skeptics among us pause in regards to the validity of their findings. A well done RCT published in JAMA by Fujii et al, examining the use of IV vitamin C, hydrocortisone and thiamine in patients with septic shock offers additional insight into the cocktail’s clinical efficacy (3).
The Vitamin C, Hydrocortisone and Thiamine in Patients With Septic Shock (VITAMINS) trial was a multicenter, open-label, parallel-group randomized trial performed in ten intensive care units in Australia, New Zealand, and Brazil. Patients who met the Sepsis-3 definition of septic shock, and were enrolled with 24-hours were randomized to receive the sepsis cocktail, IV vitamin C (1.5 g every 6 hours), hydrocortisone (50mg every 6 hours), and thiamine (200 mg every 12 hours), or hydrocortisone alone (50mg every 6 hours). Due to a lack of sufficient funds this was an open label trial. Patients in the placebo arm were not permitted to receive open label IV vitamin C.
From May 2018 to July 2019 the authors enrolled 211 patients, 107 in the intervention group and 104 in the control group. Compliance with the study protocol was excellent. The authors found no difference in their primary outcome, time alive and free of vasopressors up to day 7, (122.1 hours in the intervention group vs 124.6 hours in the control group). Nor did they find a difference in 28-day mortality, 90-day mortality, ICU mortality, hospital mortality, vasopressor free days, ventilator free days or renal replacement free days. The only outcome that demonstrated a statistical difference was the change in SOFA score at 3-days, which was slightly lower in the intervention group compared to the control group.
While some may attempt to explain the results to be due to factors other than a lack of efficacy on the part of the sepsis cocktail, it is hard to look at this study as anything other than negative. For example, the mean time to administration of the treatment in question was 12-hours. One could argue that whatever potential benefits gained from the use of vitamin C and thiamine are only seen when utilized during the initial barrage of septic shock. In addition, because of its small sample size the VITAMINS trial is incapable of conclusively ruling out the existence of a small clinical benefit. But given the lack of any signal of benefit, such arguments should be viewed as nothing more than a statistical technicalities. While the results of Fujii et al cannot eliminate the chance the sepsis cocktail provides a small survival benefit, certainly their findings disprove the spectacular efficacy initially observed by Marik et al. In addition, these negative results should reinforce the concern that the 28-day mortality benefit observed in the recently published CITRIS-ALI trial was random error due to the authors examining 42 secondary outcomes.
I have no doubt this publication will be viewed by some as a scientific I told you so. But rather than thumbing our skeptical noses at those that hoped this cocktail would provide a benefit, we should view these results as the natural progression of the scientific method. The therapeutic graveyard is full of countless interventions initially hailed as the miraculous cure for sepsis, which after a high quality RCT is performed, find little or no improvement in clinical outcomes. The pretest probability that the sepsis cocktail offered a meaningful benefit was already exquisitely low. As such, it is unlikely that the improvement in mortality observed in both the original Marik et al cohort and the CITRIS-ALI trial were due to the efficacy of vitamin C, but rather to unmeasured confounding and statistical chance. It should come as no surprise that a high quality RCT would be unable to reproduce these findings. The lack of clinical benefit observed by Fujii et al was simply the result of the scientific method taken to its most probable conclusion. The VITAMINS trial should remind us that magic bullets in medicine are exceedingly rare, especially in a complex state such as septic shock, and testing each hypothesis using an RCT, no matter how promising the preliminary data, is necessary to truly evaluate a therapy’s clinical utility.
Sources Cited:
- Marik PE, Khangoora V, Rivera R, Hooper MH, Catravas J. Hydrocortisone, Vitamin C, and Thiamine for the Treatment of Severe Sepsis and Septic Shock: A Retrospective Before-After Study. Chest. 2017;151(6):1229-1238.
- 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.
- Fujii T, Luethi N, Young PJ, et al. Effect of Vitamin C, Hydrocortisone, and Thiamine vs Hydrocortisone Alone on Time Alive and Free of Vasopressor Support Among Patients With Septic Shock: The VITAMINS Randomized Clinical Trial. JAMA.
- EM Nerd-The Case of the Partial Cohort - May 24, 2020
- EM Nerd: The Case of the Sour Remedy Continues - January 20, 2020
- EM Nerd-The Case of the Adjacent Contradictions - December 23, 2019
Just to increase uncertainty, I’ve found this publication right now in PubMed. Unfortunately the abstract is missing and I was unable to download the article:
Am J Respir Crit Care Med. 2020 Jan 9. doi: 10.1164/rccm.201908-1543LE. [Epub ahead of print]
Hydrocortisone-Ascorbic Acid-Thiamine Use Associated with Lower Mortality in Pediatric Septic Shock.
Has anyone read the full text?
Hey Rory!
As always love your writing. One caveat/reason for the negative study may be that the time to initiation of HAT therapy was significantly delayed in VITAMINS. I’m down here at EVMS with Marik and it’s certainly frowned upon if we aren’t starting HAT ASAP once we have confirmed/highly suspected sepsis or septic shock. Given what we now know from the various trials, do you think there is any possibility that timing of HAT initiation (perhaps really just the steroid) has a real effect if given early?
Cheers!
JD
Yes, look at this graph provided by Marik. In the VITAMINS study the study authors admitted that all patients getting HAT underwent an onboarding process that delayed them from getting the protocol way past the needed time, and is why Marik called it “designed to fail”. https://tinyurl.com/wkrm72g
As mentioned, time to treatment was critical, and Dr. Marik has called this a study designed to fail because the time to treatment did NOT even come close to reflecting clinical practice. One disturbing facet of this is that I see no comments or Letters to the Editor in the JAMA paper that reflect this major flaw in the VITAMINS study. Please see criticisms here, at about 33 minutes: https://www.youtube.com/watch?v=sF2ktY00dqs
As others have mentioned, the study was deeply flawed … in fact designed to fail by not giving the protocol within 6 hours of the patient demonstrating sepsis. See graph here: https://tinyurl.com/wkrm72g
Also worth noting this other study published days later that showed it cut sepsis in half:
is.gd/Xhq4YV
You can view Dr Marik’s harsh but justified indictment of the study here starting at about 33 minutes: https://www.youtube.com/watch?v=sF2ktY00dqs
I´m very interested in this topic. I´m actually working on a research project in my hospital so I can test the efficacy of Marik´s Coctail in septic shock patients. I´m even considering the chances to try it with SARS Cov 2 septic patients. There are many supporters and the trials against it lacks of statistical certainity, are open labeled or non shocked, or retrospective. I´m glad still causing interest becos many clinicians consider it old fashioned and with no future.