Editors: Kami M. Hu, MD FAAEM, Kelly Maurelus, MD FAAEM
Originally published: Common Sense
Septic shock is an illness with complex pathophysiology and few available therapies, beyond infection control and appropriate fluid resuscitation, to reverse the disease state. It is one of the most prevalent and lethal disease states that a physician may manage, with 1.7 million cases of sepsis in the United States per year and a reported mortality rate of up to 34%.1,2 The pathogenesis of septic shock is thought to be driven by a dysregulated host response3 with the role of adjunctive therapies being to assist in reversing this dysregulated response. Treatments that have more recently been a hot topic of debate include vitamin C, corticosteroids and thiamine. Vitamin C (ascorbic acid) a role in numerous physiologic processes including endothelial permeability, micro and macrovascular function, cellular apoptosis, immune system function and endogenous catecholamines.4 Studies have shown that vitamin C deficiency is present in critically ill patients,4 and its role in these essential functions is the basis for its use as a potential treatment in septic shock. Thiamine also plays a role in key metabolic processes, including cellular energy production and generation of cellular antioxidants, and thiamine deficiency has been well-documented in sepsis, with observational studies indicating a signal for improved outcomes with supplementation.4 Steroids have been used in refractory septic shock for almost the past two decades5 but the recent rationale for its use includes its synergism with vitamin C. Glucocorticoids may be able to increase the activity of vitamin C by increasing expression of the transporter involved in its uptake into cells, sodium-vitamin C transporter (SVCT2).4 In return, vitamin C, as an antioxidant, may be able to facilitate the binding of glucocorticoids to their receptor, a coupling impeded by oxidizing molecules. We will review several of the high-profile trials that have attempted to elucidate the effectiveness of utilizing corticosteroids, vitamin C, and thiamine in the management of patients with sepsis and septic shock.
- Does the treatment of sepsis with a combination of vitamin C, thiamine, and hydrocortisone improve outcomes in septic shock?
- Should steroids be used in patients with septic shock?
This retrospective before-and-after study was conducted at a single-center tertiary care referral hospital in the U.S. All patients >18 years of age with primary diagnosis of severe sepsis or septic shock and procalcitonin >2 ng/ml were enrolled. Patients were excluded if pregnant or had limitations of care. The treatment group received hydrocortisone 50mg every 6 hours for 7 days, IV vitamin C 1.5g every 6 hours for 4 days, and IV thiamine 200mg every 12 hours for 4 days. All treatments were given for the assigned time frame or until ICU discharge, whichever came first. In the control group, patients treated for severe sepsis or septic shock in the year prior to the initiation of triple therapy, hydrocortisone was used at the discretion of the intensive care unit (ICU) attending. All patients otherwise received standard treatment including broad-spectrum antibiotics, conservative fluid and vasopressor strategies, lung-protective ventilation, and appropriate prophylaxis. The primary outcome was hospital survival. Secondary outcomes included duration of vasopressor therapy, requirement for renal replacement therapy, ICU length of stay, and change in serum procalcitonin and sequential organ failure assessment (SOFA) score.
A total of 47 patients were enrolled in each arm of the study. No significant differences in baseline characteristics were reported between the two groups, though it is worth noting that p-values are not presented to support this conclusion. Hospital mortality in the treatment group was 8.5% compared to 40.4% in control group (p<0.001). Discriminant logistic analysis identified three independent predictors of mortality: APACHE IV score, need for mechanical ventilation and the vitamin C protocol (p<0.001). Three secondary outcomes were also statistically significant: duration of vasopressor therapy (18.3 h vs 54.9 h in the treatment vs control group, respectively, p<0.001), procalcitonin clearance (6.4% vs 33.9%, p<0.001), and change in SOFA score (4.8 versus 0.9, p<0.001).
These findings are impressive but exceed expected improvement and should be interpreted cautiously. Marik et al. appropriately acknowledge several major limitations to this study, including small sample size, single-center design and non-concurrent enrollment of control and treatment groups. For these reasons, it is imperative the results of this study are reproducible in a large randomized controlled trial before HAT (hydrocortisone, ascorbic acid, thiamine) therapy is accepted as standard treatment.
Fowler AA 3rd, 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. [published correction appears in JAMA. 2020 Jan 28;323(4):379].
The CITRIS-ALI Trial was a randomized, placebo-controlled, double-blinded study that was conducted in seven medical ICUs. Patients were included if they required mechanical ventilation, met criteria for acute respiratory distress syndrome (ARDS), and had a suspected or proven infection with two of four sepsis inflammatory response syndrome (SIRS) criteria, all with a 24-hour period. A total number of 167 patients were enrolled, the 84 in the treatment arm receiving 50 mg/kg of vitamin C in 5% dextrose in water (D5W) every 6 hours for 96 hours, and the 83 in the placebo arm receiving only 5% D5W every 6 hours for 96 hours.
While the trial was negative for its primary endpoints – a change in the modified SOFA (Sequential Organ Failure Assessment) score at 96 hours, and improvement in CRP or thrombomodulin levels within 168 hours – there were some secondary endpoints that reached significance. There was a reduction in 28-day mortality from 46.3% to 29.8% (p = 0.03) in the vitamin C group, an increase in ICU-free days from 7.7 to 10.7 (p = .03), and an increase in transfer out of the ICU by hour 168 from 12.5% to 25% of patients (p = .03).
Although the mortality and ICU length of stay benefits may seem compelling, there are several issues that may potentially limit the validity of the results. A large portion of screened patients (86.5%) were excluded, and the resulting sample size was perhaps too small to detect differences in primary outcomes. Reconciling a lack of difference in mSOFA scores with an improvement in mortality may be feasible if one considers survivorship bias: the sicker mSOFA scores died with the patients who did not survive. Quite appropriately, the authors point out that 46 secondary endpoints were evaluated, and a correction for multiple comparisons was not conducted, and acknowledge that these secondary results should be viewed as exploratory.
As with many bleeding-edge therapies, how practitioners view this trial will likely depend on their pre-existing beliefs. Proponents will likely point to the positive secondary outcomes of CITRIS-ALI as evidence that vitamin C improves outcomes or fall back on the argument that vitamin C is more effective when administered early in the course of sepsis, rather than after the patient has developed full-blown ARDS. On the other hand, nonbelievers are likely to point out that this was a negative trial, and it is certainly difficult to support the administration of vitamin C in septic ARDS based on this paper.
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. JAMA. 2020;323(5):423-43 .
This was a prospective, open-label, randomized controlled trial taking place across 10 ICUs in Australia, New Zealand, and Brazil. Patients were adults with presumed or confirmed infection, increase in SOFA score of at least 2 points, lactate > 2 mmol/L, and vasopressor requirement > 2 hours, enrolled within 24 hours of septic shock diagnosis. The intervention arm received 1.5g of IV ascorbate every 6 hours, 50mg IV hydrocortisone every 6 hours, and 200mg IV thiamine every 12 hours. The control arm received 50mg IV hydrocortisone every 6 hours and thiamine at the discretion of the intensivist. Treatment continued for 10 days or until the subjects maintained a MAP 65 mmHg or greater for 4 hours without the need for vasopressors.
The primary outcome was the cumulative time alive and free of vasopressors (for at least four hours) at day 7 after randomization. Predetermined secondary outcomes were mortality at 28 and 90 days, ICU and hospital mortality, 28-day cumulative vasopressor-free days, 28-day cumulative ventilator-free days, 28-day renal replacement therapy-free days, 28-day ICU-free days, hospital length of stay, and change in the SOFA score by day 3.
A total of 211 patients were included in the study: 107 in the intervention arm and 104 in the control arm. The intervention arm had lower baseline APACHE III scores but similar comorbidities, rates of mechanical ventilation and renal replacement therapy, and causes of sepsis. There was no significant difference in the primary outcome between groups (p = 0.83). There were no significant differences in the secondary outcomes between the groups with the exception of improved SOFA score at 3 days in the intervention arm (p = 0.02) a finding which must be interpreted with caution given it was applied only to patients who were in the ICU on day 3, removing from question both those who improved and left the unit and those who died within 3 days.
Though open label, the introduction of systematic performance bias may be less likely given the scale of the study, with more than 100 attending physician and intensive care fellows. The goal and achieved mean arterial pressures (MAPs) were not provided; these could affect vasopressor use. All patients received antibiotics prior to enrollment, but the time to initial antibiotic was not recorded.
Keh D, Trips E, Marx G, et al. Effect of hydrocortisone on development of shock among patients with severe sepsis: the HYPRESS randomized clinical trial. JAMA. 2016:316(17);1775-85.
The HYPRESS trial was a multicenter, placebo-controlled, double-blinded study examining the utility of hydrocortisone in preventing progression to septic shock. It included adult patients in ICUs, intermediate care units (IMCs), and community hospitals who had evidence of infection, at least 2 SIRS criteria and evidence of organ dysfunction present for < 48 hours. Exclusion criteria included septic shock (hypotension greater than 4 hours despite adequate fluid resuscitation), hydrocortisone or mannitol hypersensitivity, and history of glucocorticoid use with need for continuation of therapy. The treatment group received hydrocortisone as a dose of 50mg followed by a 24-hour infusion of 200mg for 5 days that was then tapered to 100mg on days 6 and 7, 50mg on days 8 and 9, and 25mg on days 10 and 11. The placebo group was given mannitol instead. The primary end point was the occurrence of septic shock within 14 days. Secondary end points included time to septic shock development, death, ICU and hospital mortality, vital status at days 28, 90, and 180, ICU and hospital LOS, SOFA score, duration of mechanical ventilation, renal replacement therapy requirement, and various prespecified adverse events.
A total of 380 patients were randomized, 190 to each group. There was no difference between groups in the primary outcome, and there were no significant differences in any of the secondary end points except more hyperglycemia (9.4% difference, p= 0.009) and less delirium (13.3% difference, p=0.01) in the hydrocortisone group.
Limitations include possibly missing patients who quickly developed septic shock and were therefore excluded, and the possibility that hydrocortisone may be more effective in patients who are more severely ill. The major limitation, however, is that the study population’s lower-than-anticipated incidence of septic shock (23% vs the expected 40%) resulted in an underpowering of the trial and inconclusive results.
Venkatesh B, Finfer S, Cohen J, et al. Adjunctive glucocorticoid therapy in patients with septic shock. New Engl J Med. 2018;378(9):797-808.
The ADRENAL Trial was an investigator-initiated, international, pragmatic, double-blind, parallel-group, randomized, controlled trial that compared IV infusions of hydrocortisone with matched placebo in adult patients with septic shock on vasopressors who were undergoing mechanical ventilation in an ICU. Patients in the experimental arm of the study were given an intravenous infusion of hydrocortisone at a fixed dose of 200 mg per day. Both experimental and control groups were well matched with respect to age, sex, site of infection, and type of ICU. The primary outcome of the study was all-cause mortality at 90 days. Secondary outcomes included 28-day mortality, time to shock resolution, ICU and hospital length of stay (LOS), mechanical ventilation and renal replacement therapy requirement and duration, incidence of subsequent bacteremia/fungemia, and blood transfusion requirement.
A total of 3658 patients were enrolled: 1832 patients randomized to the hydrocortisone group and 1826 to the control group. At 90 days, there was no significant difference in mortality [27.9% vs 28.8%, (odds ratio, 0.95; 95% confidence interval [CI], 0.82 to 1.10; P=0.50)]. The treatment group had faster shock resolution [3 vs 4 days (hazard ratio [HR] 1.32, 95% CI 1.23-1.41)], more ventilator-free days [6 vs 7 days (HR 1.13, 95% CI 1.05-1.22, p<0.001)], and a shorter ICU LOS [(10 vs 12 days (HR 1.14, 95% CI 1.06-1.23, p<0.001)]. Also, fewer patients in the treatment group received a blood transfusion (37.0% vs. 41.7%; odds ratio, 0.82; 95% CI, 0.72 to 0.94; P=0.004), without a significant between-group difference in mean total volume of blood transfused when examining all patients who received transfusions. There were more adverse events (hyperglycemia, hypernatremia, encephalopathy, myopathy, etcetera) in the hydrocortisone group (1.1 vs 0.3%, p=0.009).
Investigators did not clarify if additional sepsis management was appropriate (i.e., correct antibiotics, appropriate fluid resuscitation), so the potential for confounding remains. They excluded patients who received etomidate, a medication that remains widely used for rapid sequence induction, which limits generalizability of the results, and did not assess for actual adrenal insufficiency. Finally, it remains unclear whether continuous infusion of hydrocortisone is less beneficial than bolus dosing due to slower drug delivery.
Ultimately, the ADRENAL Trial seems to indicate that infusions of hydrocortisone did not improve 90-day mortality in septic shock. It does support consistent findings of earlier shock resolution and ventilator weaning, and in general is unlikely that this paper will change independent physicians’ practice given its limitations, which could be argued for or against its validity.
Annane D, Renault A, Brun-Buisson C, et al. Hydrocortisone plus Fludrocortisone for Adults with Septic Shock. New Engl J Med. 2018;378(9): 809-18.
The AProCCHSS trial (Activated Protein C and Corticosteroids for Human Septic Shock), so named because it was initially designed as a 2x2 factorial study investigating hydrocortisone plus fludrocortisone versus droctrecogin alfa versus placebo, was continued as a multicenter randomized controlled trial without droctrecogin after it was pulled from the market.
A total of 1241 patients were enrolled from 34 different centers, with randomization occurring in a permuted block of eight patients. Major inclusion criteria included an indisputable or probable diagnosis of septic shock, a SOFA score of at least 3-4 in at least two different organs, and a vasopressor requirement for at least 6 hours at a dose of at least 0.25 micrograms per kg per minute (mcg/kg/min) of norepinephrine (or equivalent-dose alternative vasopressor). Major exclusion criteria included presence of septic shock for >24 hours since transfer to the ICU, a high risk of bleeding, pregnancy/lactation, or prior treatment with corticosteroids.
The treatment group received hydrocortisone 50 mg every 6 hours and fludrocortisone 50 mcg daily for 7 days without taper, while the placebo group received mannitol and microcrystalline cellulose. In addition, attempts were made to harmonize non-experimental interventions between centers including anti-infective treatments, hemodynamic and respiratory management, blood glucose control, and neuromuscular blockade use. A steering committee also judged the adequacy of the anti-infective coverage which was found to be similarly adequate in both the treatment and placebo group (96.9% vs 96.2%).
Patient characteristics were similar between both intervention and placebo groups including an average age of 66 years and male predominance (65.5% vs 67.7%. The majority were admitted as an escalation from the medical wards with pulmonary and urinary infections being the most common. Norepinephrine was the most common vasopressor used. The primary outcome was 90-day all-cause mortality which was found to be significantly lower in the steroid group [43.0% vs 49.1%; RR (relative risk) 0.88 (95% CI 0.78 – 0.99, p=0.03)]. The steroid group also had a decrease in both death from any cause at ICU (35.4% vs 41.0%, p=0.04) and death by day 180 (46.6% vs 52.5%, p=0.04, with significantly more vasopressor-free days through day 28 (17 vs 15 days , p<0.001). With the exception of increased hyperglycemia incidence in the treatment group, rates of adverse events (serious bleeding, superinfections, neurologic sequelae) were similar.
The authors argue two significant differences between the steroid trials that have shown mortality benefit and those that did not: (1) the use of the mineralocorticoid fludrocortisone, which may have contributed to intravascular volume expansion, and (2) the requirement for at least 6 hours of vasopressor therapy, likely selecting for sicker patients who do not improve with the standard 6-hour Surviving Sepsis Campaign bundle of care. They argue that the inclusion and exclusion criteria for this study select for a sicker patient population with worse illness severity scores, which makes them an ideal group for adjunctive therapies. With the confidence interval for mortality reduction approaching 1.0, the benefit should be considered potential and needs replication in additional studies.
The Marik trial was certainly attention-grabbing and left many feeling hopeful that an effective therapy for septic shock with significant mortality benefit had been discovered, despite its various limitations. The CITRIS-ALI trial showed no benefit in the non-patient-oriented primary outcome of reduction in SOFA scores, CRP or thrombomodulin. Its 28-day mortality benefit, as a secondary outcome in a study that did not adjust for multiple comparisons, can be nothing more than hypothesis-generating. The VITAMINS trial showed no benefit to adding vitamin C and thiamine to the widely-used hydrocortisone. Ongoing studies including VICTAS (vitamin C, Thiamine, and Steroids in Sepsis),6 ACTS (Ascorbic acid, Corticosteroids, and Thiamine in Sepsis)7 and others will be needed before the universal use of vitamin C, thiamine and corticosteroids is implemented in patients with septic shock.
The controversy regarding the use of steroids has been ongoing for over a decade. Both Annane’s APROCCHSS trial and his trial in 2002 showed a mortality benefit to hydrocortisone and fludrocortisone,12 while the results of the ADRENAL trial mirror those of the CORTICUS trial: no difference in mortality with faster shock resolution.8 It is important to note that the APROCCHSS trial did have patients with higher overall mortality than the ADRENAL and CORTICUS trials, suggesting that the benefit from steroids may only be seen in the sickest of patients. What is most consistent is that several of these studies have demonstrated faster shock resolution, ventilator weaning, and/or shortened ICU LOS, with slight increase in hypernatremia and hyperglycemia. While steroids for all septic patients does not seem supported, in a patient who is critically ill and not responding to antibiotics, appropriate fluids, and high-dose and/or multiple vasopressors, it seems reasonable to add steroid therapy to ongoing life-saving efforts.
- Existing evidence does not clearly support improved clinical outcomes with the combination of vitamin C, thiamine, and hydrocortisone in septic shock.
- For septic shock refractory to standard therapies, steroids hasten shock resolution and are probably worth initiating, but their effect on mortality remains unclear.
- Rhee C, Dantes R, Epstein L, et al. Incidence and trends of sepsis in US hospitals using clinical vs claims data, 2009-2014. JAMA. 2017;318(13):1241-9.
- Hatfield K, Dantes R, Baggs J, et al. Assessing Variability in Hospital-Level Mortality Among U.S. Medicare Beneficiaries With Hospitalizations for Severe Sepsis and Septic Shock. Crit Care Med. 2018; 46(11):1753-60.
- Angus D, van der Poll T. Severe Sepsis and Septic Shock. New Engl J Med. 2013;369(9):840-51.
- Moskowitz A, Andersen L, Huang D, et al. Ascorbic acid, corticosteroids, and thiamine in sepsis: a review of the biologic rationale and the present state of clinical evaluation. Critical Care. 2018;22(1): 283.
- Annane D, Sébille V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA. 2002;288(7):862871.
- Hager D, Hooper M, Bernard G, et al. The Vitamin C, Thiamine and Steroids in Sepsis (VICTAS) Protocol: A prospective, multi-center, double-blind, adaptive sample size, randomized, placebo-controlled, clinical trial. Trials. 2019. doi: 10.1186/s13063-019-3254-2.
- Moskowitz A, Yankama T, Andersen LW, et al. Ascorbic Acid, Corticosteroids and Thiamine in Sepsis (ACTS) protocol and statistical analysis plan: a prospective, multicentre, double-blind, randomised, placebo-controlled clinical trial. BMJ Open. 2019;9(12):e034406. doi:10.1136/bmjopen-2019-034406.
- Sprung CL, Annane D, Keh D, et al. Hydrocortisone Therapy for Patients with Septic Shock. New Engl J Med. 2008;358(2):111-24.