Phase I safety trial of intravenous ascorbic acid in patients with severe sepsis

The incidence of sepsis and sepsis-associated organ failure continues to rise in Intensive Care Units worldwide with studies from multiple countries showing that organ failure contributes cumulatively to patient mortality [1‐3]. Patients with severe sepsis suffer higher mortality rates compared to patients with organ failure but no sepsis. Despite over 15,000 patients studied and over 1 billion dollars in study costs effective sepsis therapy remains elusive [4, 5]. Clinical trials that have targeted mediators of inflammation or coagulation such as atorvastatin [6] or activated protein C [7] have not reduced septic mortality, suggesting that single-target therapy fails to meet the challenges of complex multicellular activation and interactions.

Recent studies suggest that ascorbic acid may attenuate pathological responses in septic microvasculature. Armour et al. and Wu et al. showed that ascorbic acid infusion improved capillary blood flow, microvascular barrier function, and arteriolar responsiveness to vasoconstrictors in septic animals [8, 9]. Recently, we showed that parenterally infusing ascorbic acid at a concentration of 200 mg/kg attenuated vascular lung injury in septic mice by multiple mechanisms, including attenuation of the proinflammatory mediators, enhanced alveolar epithelial barrier function, increased alveolar fluid clearance, and prevention of sepsis-induced coagulopathy [10, 11]. In addition, ascorbic acid deficient mice were found to be more susceptible to sepsis-induced multiple organ dysfunction and parenteral infusion of ascorbic acid attenuated the injury (lung, kidney, liver) [12].

Subnormal plasma ascorbic acid concentrations in septic patients correlate inversely with the incidence of multiple organ failure and directly with survival [13]. Ascorbic acid depletion in sepsis results from: 1) ascorbic acid consumption by reduction of plasma free iron, 2) ascorbic acid consumption by the scavenging of aqueous free radicals, and 3) by destruction of the oxidized form of ascorbic acid, dehydroascorbic acid [14]. Dosing and bio-distribution data in humans show that pharmacological concentrations of ascorbic acid can only be attained following intravenous administration [15]. Surprisingly, few studies in critically ill patients infusing ascorbic acid have been performed. Nathens and colleagues infused ascorbic acid at 1 gram every 8 hours combined with oral vitamin E for 28 days in 594 surgically critically ill patients and found a significantly lower incidence of acute lung injury and multiple organ failure [16]. Tanaka et al. infused ascorbic acid continuously at 66 mg/kg/hour for the first 24 hours in patients with greater than 50% surface area burns and showed significantly reduced burn capillary permeability [17]. A single report (published as abstract only) of a clinical study of large intravenous doses of ascorbic acid, and other antioxidants (tocopherol, N-acetyl-cysteine, selenium), in patients with established ARDS showed a 50% reduction in mortality [18]. Clinical protocols currently in use for hospitalized septic patients fail to normalize ascorbic acid levels. Ascorbic acid dosages utilized in this trial arose from our preclinical work.

In the current trial, we sought to determine whether intravenous ascorbic acid was safe to administer to critically ill patients with severe sepsis and to determine if ascorbic acid had an impact on organ failure and a priori selected blood biomarkers. We measured C-reactive protein and procalcitonin as systemic markers of inflammation while choosing thrombomodulin as a marker of vascular injury [19‐21]. The work reported in this study has previously been presented at the American Thoracic Society International Meeting [22].

This study was approved by the VCU Institutional Review Board (IRB). The IRB approval number assigned to this trial was: HM12903. The trial was conducted under a randomized double blind placebo-controlled format. A multi-departmental data safety monitoring board oversaw the trial.

This phase I trial focused on the safety of administering intravenous ascorbic acid to patients with severe sepsis. The intravenous route of administration was chosen in this trial in order to achieve high ascorbic acid plasma levels. Padayatty and colleagues showed that high-level ascorbic acid plasma concentrations could only be achieved by intravenous administration [15]. Prior human studies employing pharmacologic ascorbic acid dosing report no adverse events. Nathens et al. administered 1 gram of ascorbic acid every 8 hours for 28 days to surgically critically ill patients with no ill effects [16]. Tanaka et al. administered 66 mg/kg/hour for 24 hours to patients with 50% surface area burns with no adverse events [17]. Hoffer et al. intravenously administered up to 90 grams of ascorbic acid 3 times weekly to patients with advanced malignancy with no adverse events [30]. The dosing protocols we chose for this trial arose out of our preclinical work.

Prior studies show that patients with severe sepsis exhibit significantly reduced plasma ascorbic acid levels upon admission to intensive care [31]. The mean initial plasma ascorbic acid level for all septic patients in this study was 17.9 ± 2.4 μM compared to normal human plasma levels of 50 – 70 μM (Figure 1). Prior studies [13, 14, 26], and the current study show that subnormal plasma ascorbic acid levels are a predictable feature in patients with severe sepsis. Importantly, Placebo patients exhibited no change in plasma ascorbic acid levels throughout the 4-day study period despite receiving full ICU standard of care practice for severe sepsis (Figure 1). Ascorbic acid depletion in sepsis results from ascorbic acid consumption by the reduction of plasma free iron, ascorbic acid consumption by the scavenging of aqueous free radicals (peroxyl radicals), and by the destruction of the oxidized form of ascorbic acid dehydroascorbic acid[14]. Sepsis further inhibits intracellular reduction of dehydroascorbic acid, producing acute intracellular ascorbic acid depletion. Sepsis-induced ascorbic acid destruction permits uncontrolled oxidant activity which amplifies tissue injury [14, 32, 33]. Ascorbic acid treated patients in this study exhibited rapid and sustained increases in plasma ascorbic acid levels using an intermittent every six hours administration protocol (Figure 1).

SOFA scores are robust indicators of mortality during critical illness. SOFA score increases during the first 48 hours of ICU care predict a mortality rate of at least 50% [26]. In this study, the extent of organ failure accompanying patients with severe sepsis was high with an average SOFA score for all patients equal to 11.4 ± 3 confirming that multiple organ failure was present at enrollment. Given that the mean plasma ascorbic acid levels on admission were subnormal (17.9 ± 2.4 μM), a mean initial SOFA score of 11.4 ± 3 in patients with severe sepsis was not surprising. This study is in agreement with other studies which show that plasma ascorbic acid levels in severe sepsis correlate inversely with the incidence of multiple organ failure [13]. We showed that the addition of ascorbic acid to standard of care practice (i.e., fluid resuscitation, antibiotics, vasopressor medication) for patients with severe sepsis significantly reduced organ injury. Ascorbic acid treated patients exhibited prompt and sustained reductions in SOFA scores during the 4-day treatment regimen unlike placebo controls where SOFA scores slowly increased over time. SOFA score reduction was most remarkable in patients receiving the high dose ascorbic acid infusion (Figure 2).

C-reactive protein (CRP) [19] and procalcitonin (PCT) [20] levels are known to correlate with the overall extent of infection and higher levels of both have both been linked to higher incidences of organ injury and death in the critically ill. CRP in circulation has a short half-life of approximately 19 hours. Thus, the kinetics of CRP make it a useful monitor for tracking the inflammatory response produced by infection, and the response to antibiotic treatment. Lobo et al. reported that patients with CRP levels greater than 10 mg/dL at ICU admission exhibited significantly higher rates of multiple organ failure as well as higher mortality rates [34]. A decrease in CRP levels in Lobo’s study after 48 hours was associated with a mortality rate of only 15.4%, while a persistently high CRP level was associated with a mortality rate of 60.9%. Both low and high dose ascorbic acid infusion in this trial promptly reduced serum CRP levels in septic patients (Figure 3A). Thus, the findings in this study support the findings of Lobo et al. with descending CRP levels being associated with lower mortality rate and reduced levels or organ failure. Jensen and colleagues found that high maximal procalcitonin levels were an early independent predictor of all-cause mortality in a 90-day follow-up period after intensive care unit admission [20]. Karlsson and colleagues [21] showed that mortality in patients with severe sepsis was lower in those patients in whom procalcitonin concentrations fell by more than 50% at 72 hours with respect to initial values. Infusion of ascorbic acid into patients with severe sepsis in this study reduced serum procalcitonin levels by greater than 50% (Figure 3B). Thrombomodulin is an endothelial cell bound molecule that captures thrombin holding it adjacent to protein C bound to its receptor (endothelial protein C receptor). Elevated soluble TM in the circulation indicates endothelial cell injury [35]. Lin et al. reported that increased TM levels correlated with the extent of organ failure and mortality in patients with sepsis [36]. In the current study, thrombomodulin levels in patients randomized to placebo began to rise at approximately 36 hours into the study period, indicating sepsis-induced endothelial injury (Figure 4). Patients randomized to receive either dose of ascorbic acid exhibited no subsequent rise in plasma thrombomodulin. Though our patient numbers were small, these early results suggest that intravenous ascorbic acid acts to attenuate the proinflammatory state of sepsis and perhaps attenuates the development of endothelial injury.

On the basis of this study and our prior preclinical studies, we speculate as to the pleiotropic mechanisms by which ascorbic acid would be beneficial in sepsis. Ascorbic acid is rapidly taken up by endothelial cells in millimolar quantities where it scavenges reactive oxygen species and increases endothelial nitric oxide synthase-derived nitric oxide by restoring tetrahydrobiopterin content, thus, increasing bioavailable nitric oxide. As we and others have shown in basic investigations [10, 11, 37], by inhibiting NFκB activation, ascorbic acid could potentially attenuate the “cytokine storm” that arises due to NFκB driven genes known to be activated in sepsis. Septic ascorbic acid-deficient neutrophils fail to undergo normal apoptosis. Rather, they undergo necrosis thereby releasing hydrolytic enzymes in tissue beds, thus contributing to organ injury. We speculate that intravenous ascorbic acid acts to restore neutrophil ascorbic acid levels. Repletion of ascorbic acid in this way allows for normal apoptosis, thus, preventing the release of organ damaging hydrolytic enzymes. A multitude of biological mechanisms are active in patients with sepsis and they promote multiple organ injury and death.

Tens of thousands of lives are lost across the world annually due to severe sepsis [1, 38‐40]. Multiple treatment trials have failed to measurably improve outcomes. The majority of trials have singly eliminated certain proinflammatory mediators which research has suggested promotes tissue damage. The single mediator approach has largely been unsuccessful. The results from this small phase I safety trial suggest that administering ascorbic acid in pharmacological dosages to critically ill patients with sepsis is safe and that it may provide adjunctive therapy in the treatment of severe sepsis. A larger phase II proof-of-concept trial is needed.

This phase I trial shows that aggressive repletion of plasma ascorbic acid levels in patients with severe sepsis is safe. This early work in septic patients suggests that pharmacologic ascorbic acid repletion reduces the extent of multiple organ failure and attenuates circulating injury biomarker levels.