Is a vitamin C–based cocktail the cure for severe sepsis and septic shock? If you read headlines and news items that appeared widely in the mainstream media in 2017, you might have concluded that it is. After all, a recent paper in Chest had found that when a single intensive care unit (ICU) rolled out a new protocol in which a “cocktail” of vitamin C, thiamine, and hydrocortisone was given to patients with severe sepsis or septic shock, mortality fell from 40.4 percent to 8.5 percent.¹ This result was both extraordinary and almost implausible, despite some compelling physiological justifications bolstering the theory.

The hype could barely be controlled. Some physicians began using the cocktail right away. A flurry of trials were designed and approved by hospital review boards around the globe. Trialists moved as quickly as they could to begin studying this seriously and employing a variety of rigorous methodologies, assessing various patient populations and a variety of outcome measures. We needed some answers, and we needed them quickly.

Last month, the results from the Vitamin C, Hydrocortisone and Thiamine in Patients With Septic Shock (VITAMINS) trial, the first major international multicenter randomized, controlled effort to be completed, were unveiled in JAMA.2

The findings: negative. Across the board.

The Results

For virtually every outcome that the authors assessed for efficacy in septic shock, the patients who received the vitamin C–based cocktail (often referred to as the Marik protocol, metabolic resuscitation, or HAT for hydrocortisone, ascorbic acid, and thiamine) experienced no added benefit over patients in the control arm who received hydrocortisone only.

The study’s primary outcome was duration of time alive and free of vasopressor administration up to day 7 of treatment. The trial, which enrolled patients from 10 hospitals in Australia, New Zealand, and Brazil, also reported data on 10 prespecified secondary outcomes, including 28- and 90-day mortality, the need for dialysis, and mechanical ventilation, among others. For each of these, the Marik protocol failed to bestow any benefit. Out of 10 secondary outcomes, the only signal of benefit to emerge from the VITAMINS trial was a one-point improvement over controls in the sequential organ failure assessment score. However, as the other outcomes clearly demonstrate, patients who received the Marik protocol fared no better overall in any patient-centered outcome. Numerically, though not statistically, more deaths actually occurred in the vitamin C cocktail group. (In fairness, this was not even a trend; it is only worth mentioning because so very few patients died in the vitamin C group in the 2017 study.)

This will be seen as a major disappointment for observers desperately looking for therapies to offer patients with life-threatening sepsis syndromes. However, this was always the most likely outcome, given the improbable magnitude of reported benefit found in the original Marik study, upon which the entire vitamin C frenzy has been based. Alas, desperation, hope, and hype were never going to be enough. What we always needed was a well-executed trial to either confirm or refute Marik’s hypothesis and potentially game-changing findings. We now have the first credible report. From the looks of it, it’s back to the drawing board.

Breaking Down What Happened

How did we get here? When the vitamin C protocol first made news in 2017, there were two polar responses to this. The believers celebrated the treatment as a brilliant innovation, based on a genuine understanding of complex physiology, that could save hundreds of thousands of lives. The skeptics pointed out that the study upon which the excitement was based was a before-after retrospective chart study design performed in a single ICU. Many noted that the study, designed and led by the outspoken intensive care physician Paul Marik, MD, FCCP, was not a genuine trial. Rather, it amounted to a quality improvement project. In the Marik study, all patients with severe sepsis or septic shock received the vitamin C–based cocktail over a six-month period in early 2016. Those outcomes were then compared to those of a similar number of patients treated in that ICU before the protocol was rolled out in 2016 and who met similar inclusion criteria.

It bears mentioning that quality improvement studies almost always yield favorable results for the problem being addressed. When resources—institutional, financial, and cognitive—are being applied to a challenging task, the short-term results are frequently good, yet difficult to maintain. The hidden costs of quality improvement efforts, however, are difficult to assess and not usually reported. For example, if a project to expedite CT for every patient with a neurological complaint is undertaken, a few patients may have their cerebrovascular accidents diagnosed sooner. But how many patients with acute aortic syndromes had diagnoses delayed because of that? Similarly, when we concentrate substantial human and financial resources onto one problem, do other problems suffer in silence?

This is why randomized, controlled trials are required to make statements about the effectiveness of experimental therapies. Double-blinding is preferred as well because patients in the control arm are assured to receive as much attention as those in the intervention arm. The VITAMINS trial was unblinded (open label). This occurred because the study was initially an unfunded “passion” project by the team of investigators. Blinding is expensive and requires administrative muscle (which is not free). Treatments need to be concealed from both the providers and the subjects. While this study would be stronger if it had been blinded, you could argue that the lack of blinding favored the intervention. While we can’t know whether the authors were skeptical or optimistic about the cocktail’s chances, it is difficult to imagine that providers in 10 ICUs across three countries were all biased against an inexpensive therapy that had the potential to save lives.

As before, the mainstream media covered this story. National Public Radio, which widely publicized the protocol in 2017, again took notice. Dr. Marik told NPR that “in his experience, the treatment is only effective if given within six hours after someone has suspected sepsis.” (The typical time-to-treatment with the cocktail was around 12 hours in the VITAMINS trial.) But this is simply another way of saying that early detection and treatment of severe sepsis and shock are important. The patients for whom Dr. Marik declares the protocol is effective are precisely the ones receiving timely treatments we know to be crucial, including antibiotics and, in some patients, fluids and vasopressors. It is safe to propose that for patients who do not receive these proven therapies promptly, nothing will work later. However, in the VITAMINS trial, that is not what occurred. Instead, all patients received antibiotics prior to randomization. Nor do we know how quickly patients received either antibiotics or the vitamin C cocktail in the Marik study, as these data were not reported. However, the authors of the VITAMINS trial have already indicated that a subgroup analysis that takes time-to-treatment into account may be forthcoming.

Eighteen other studies assessing this cocktail are under way. With that many trials, each with its own patient inclusion criteria and unique outcome measurements, one of them is bound to find some signal of benefit by chance alone. But based on the VITAMINS study, I believe we can conclude that this miracle cure is not to be. If benefit is uncovered by any of these subsequent trials, it is likely to be small and incremental at best. Knowing that, we must again widen our perspective in our continued search for therapies that can truly turn the tide against sepsis.

References

1. Marik PE, Khangoora V, Rivera R, et al. 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.
2. 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 [published online ahead of print Jan. 17, 2020]. JAMA. doi:10.1001/jama.2019.22176.