During this epoch of COVID-19, it might be nice to take a little breather and have an angry argument about a different topic – vitamin C! Where were we? The last chapter of this story was the VITAMINS trial, which found no benefit from vitamin C.1 However, critics of this trial suggested that patients were enrolled too late to see benefit, so perhaps Vitamin C would have worked better if started earlier. Very well, what’s new?
Chang et al. Combined treatment with hydrocortisone, vitamin C, and thiamine for sepsis and septic shock (HYVCTTSSS trial)
This is a single-center RCT evaluating the combination of hydrocortisone, vitamin C, and thiamine for patients with sepsis or septic shock.2 The study was intended to be single-blinded, but it seems that the blinding scheme didn’t really work (in many cases treating teams stopped bothering to give the placebo infusion).
The original design was to recruit 140 patients, but the study was stopped prematurely after recruiting only 80 patients due to concerns that the treatment was causing hypernatremia (sodium >160 mM occurred in 13 patients in the treatment arm and 3 in the control arm). This appears to be a violation of the study’s stopping rule (which specified that the trial should be stopped only if an adverse event occurred with p < 0.005). Using a Fisher Exact test, the p-value of hypernatremia is p=0.01, which is above the cutoff to stop the trial:
There are additional reasons that stopping the study due to hypernatremia was silly. Hydrocortisone is known to promote hypernatremia, but this is a very mild effect which is easily managed with usual high-quality ICU care (e.g., daily monitoring of electrolytes and aggressive administration of free water to treat any hypernatremia – as should usually be provided to any ICU patient). If patients’ sodium values were being allowed to gradually drift over 160 mM, this is more a reflection of the quality of the ICU care overall than of any specific intervention.
Patients were included if they met the definition of sepsis-3 and had a procalcitonin >2 ng/mL. At baseline, patients were fairly well matched:
The primary endpoint was mortality, with a trend towards reduced mortality in patients treated with vitamin C (27.5% vs. 35%, p=0.47). However, because of under-powering it is impossible to make heads or tails of this result. The 95% confidence interval for the relative risk of mortality was 0.79, with a 95% confidence interval from 0.41 to 1.52. This enormously wide confidence interval indicates that the study fails to evaluate for significant benefits or harms.
To illustrate how ridiculously under-powered this trial is, I’ve calculated the relative risk reduction from vitamin C in this trial and in Marik’s original trial, using a mortality endpoint (table below).3 For the sake of statistical uniformity, both calculations were performed using this online calculator. As shown below, the 95% confidence interval for the relative risk from both studies overlaps (e.g., both studies are consistent with a relative risk of 0.5 from vitamin C therapy). This illustrates that the current study neither contradicts nor disproves Marik’s prior study – it’s too underpowered to do anything.
The secondary endpoints are fairly ambiguous (table below). There was a small reduction in vasopressor duration, but underpowering prevents this from coming close to statistical significance. Improvement in SOFA score was higher in the treatment group, but as an isolated secondary endpoint it’s hard to know what to make of this.
Many patients in the study were referred from other centers, raising the possibility that intervention occurred too late in the disease course to be effective. In a pre-specified subgroup analysis, mortality benefit was observed in patients diagnosed with sepsis within <48 hours of ICU admission (figure below). This could be consistent with the concept that vitamin C requires earlier administration to work. However, this is statistically fragile and not supported by many signals of benefit elsewhere in the study.
So overall the study is woefully underpowered. If we can overlook the silliness with the hypernatremia, there are some possible signals of benefit – but they are rather dim. Perhaps most importantly, though, it’s impossible to sort out benefit due to vitamin C versus benefit due to steroid. Corticosteroid is known to improve hemodynamic stability in septic shock.4,5 Thus, it’s to be expected that the cocktail would cause some benefit due solely to the steroid component. The key question is whether vitamin C and thiamine add anything on top of the steroid – a question which isn’t answerable from this study.
- HYVCTTSSS trial is a single-center RCT evaluating combination therapy with hydrocortisone, ascorbate, and thiamine for patients with sepsis or septic shock.
- The trial was prematurely terminated due to difficulty in managing hypernatremia, leaving the study underpowered.
- The primary endpoint was mortality. Hydrocortisone, ascorbate, and thiamine caused a relative risk of mortality with a 95% confidence interval ranging from 0.41 to 1.52. This intervention might therefore be extremely effective (RR 0.41), or highly dangerous (RR 1.52), or totally ineffective (RR 1.0).
- Some secondary endpoints did suggest possible benefit. However, it’s impossible to sort out potential benefit from hydrocortisone alone versus the combination of ascorbate, hydrocortisone, and thiamine.
- Overall this trial provides no meaningful information about hydrocortisone, ascorbate, and thiamine.
- 1.Fujii T, Luethi N, Young P, 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. January 2020. doi:10.1001/jama.2019.22176
- 2.Chang P, Liao Y, Guan J, et al. Combined treatment with hydrocortisone, vitamin C, and thiamine for sepsis and septic shock (HYVCTTSSS): A randomized controlled clinical trial. Chest. March 2020. doi:10.1016/j.chest.2020.02.065
- 3.Marik P, Khangoora V, Rivera R, Hooper M, 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. doi:10.1016/j.chest.2016.11.036
- 4.Venkatesh B, Finfer S, Cohen J, et al. Adjunctive Glucocorticoid Therapy in Patients with Septic Shock. N Engl J Med. 2018;378(9):797-808. doi:10.1056/NEJMoa1705835
- 5.Annane D, Renault A, Brun-Buisson C, et al. Hydrocortisone plus Fludrocortisone for Adults with Septic Shock. N Engl J Med. 2018;378(9):809-818. doi:10.1056/NEJMoa1705716