Severe sepsis is a challenging problem in the emergency department or intensive care unit (ICU) 1, and can lead to septic shock or multiple organ failure. The complex mechanisms underlying severe sepsis remain unclear. In sepsis, the overwhelming inflammatory response to the invading pathogen is the major pathophysiologic challenge, rather than the pathogen itself. In a systemic inflammatory response, both endothelial cells and neutrophils are activated to release oxygen-derived free radicals 2. It seems that these oxyradicals play a role in causing or propagating the systemic inflammatory response syndrome (SIRS) in life-threatening conditions, and that the imbalance in redox state reflects both oxidative stress and tissue damage 34.

Measurement of serum total antioxidant capacity (TAC) level was reported to provide an integrated index, as opposed to one based on simple summation of measurable antioxidants 5. It possibly could be used to assess the real change in antioxidant status in patients with severe sepsis and might lead to universally useful treatment 6. Several preclinical and clinical studies of sepsis focused on single-point inhibition (for example, anti-tumor necrosis factor antibodies) or augmentation of specific key processes, and failed to demonstrate therapeutic efficacy 7. Most believe that higher levels of oxyradicals and lower antioxidant levels in patients with SIRS or septic shock lead to multiple organ failure 8910. However, serum TAC increases in critically surgical patients with septic shock 11. Moreover, endogenous peroxyl-radical scavenging ability in the plasma of SIRS patients was found to be elevated in nonsurvivors 12. The actual change in TAC in severe sepsis remains controversial.

Serum uric acid (UA), like other antioxidants such as albumin, bilirubin, or vitamins A, C and E, is a powerful free-radical scavenger and increases in response to acute oxidative stress 1113. UA formation may even provide a significant antioxidant defense mechanism against nitration by peroxynitrite in rat heart during hypoxia 14. It is therefore postulated that serum UA level is an important marker in oxidative stress. Recently, serum UA was identified as a strong predictor of mortality in patients with moderate-to-severe chronic heart failure 15. This finding raises an interesting question about the actual pathophysiologic role of serum UA in critically ill patients.

We conducted the present study to investigate whether serum TAC levels are elevated or suppressed in emergency department patients with severe sepsis. We also wished to determine the correlation between serum TAC level and severity of illness, and the relationship between serum TAC and UA levels in emergency department patients with severe sepsis.

The study was conducted between April 2001 and March 2003 at an academic tertiary care center, which receives between 43,000 and 54,000 emergency department visits each year. The study was approved by our hospital's Institutional Research Review Board.

The study enrolled 149 participants aged over 18 years: 73 patients who met the criteria for severe sepsis (including septic shock), as defined by the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference; and 76 healthy control individuals. These patients (41 men and 32 women; mean age 65.9 ± 16.4 years) presented with a mean APACHE II score of 23.2 ± 8.2; the mortality rate was 26.0% (19 out of 73; Table 1). Of the patients, 39 (53.4%) had pneumonia, nine (12.3%) had soft tissue infections (including necrotizing fasciitis and Fournier gangrene, among others), seven (9.6%) had urinary tract infections, six (8.2%) had biliary tract infections, three (4.1%) had central nervous system infections (including brain abscess, meningitis, among others), and nine (12.3%) had unknown foci of infection. Nonsurvivors had higher APACHE II scores and a higher ratio of septic shock in the emergency department than did survivors. The leading infectious micro-organism were Gram-negative bacteria (35.6%) and the positive culture rate was 75.3% (55 out of 73). Detailed demographic data, clinical diagnoses, and microbiological data for patients and healthy control individuals are summarized in Table 1.

In this study we showed that serum TAC levels reflected clinical severity of sepsis. Serum TAC levels became elevated as APACHE II scores rose. In addition, serum TAC levels were higher in nonsurvivors than in survivors. Why serum TAC increased in patients with severe sepsis is unknown. It is hypothesized that anti-inflammatory processes are activated to counterbalance excessive levels of proinflammatory cytokines 22 or oxidative stress 23. It is also possible that altered antioxidant defense (i.e. a significant change in serum TAC level) eventually leads to immune dysfunction and a poor outcome 111224.

Higher sustained serum TAC levels, as noted in the nonsurvivors, might be a host response to severely propagating oxidative stress or a compensating mechanism for depleted antioxidative components 13. It is also possible that the increase in serum TAC simply worsened severe sepsis, which suggests that antioxidant therapy should not routinely be used in the treatment of severe sepsis until its mechanism of action is understood.

TAC levels in patients with severe sepsis remain controversial. Although Ghiselli and coworkers 5 suggested that plasma rather than serum should be used to measure TAC levels, in the study by Whitehead colleagues 17 and in ours 18 serum was used. In one of our other experiments there appeared to be no significant difference between the levels detected in plasma and serum (data not shown). Pascual and coworkers 11 reported that plasma TAC levels were lower in patients with sepsis but higher in patients with septic shock than in control individuals. MacKinnon and colleagues 13, in a study conducted in 50 critically ill patients, reported that total antioxidant status and UA levels were higher in nonsurvivors than in survivors, and speculated that the higher total antioxidant status level might have reflected the higher UA levels caused by renal dysfunction. In accordance with our findings, Tsai and coworkers 12 reported that plasma TAC levels were significantly higher in nonsurvivors than in survivors. However, most other studies 2825 found that the plasma antioxidant potentials of nonsurvivors were significantly lower than those of survivors among patients with severe sepsis.

The discrepancy in the findings from these studies may mainly be due to differences in measurement methods. Instead of using a spectrophotometric method 2, we and two other groups 1112 used the TRAP method to measure the total antioxidant potential. The spectrophotometric method, proposed by Miller and coworkers 26, measures the reaction of plasma antioxidants and a mixture of different radical species including 2,2'-azinobis(3-ethylbenzothiazoline 6-sulfonic acid; ABTS) radical cation, ferryl myoglobin, hydrogen peroxide, and other radicals. In the spectrophotometric methods, albumin and UA account for 43% and 33%, respectively, of TAC. The marked decrease in serum albumin in patients with severe sepsis may lead to lower levels of TAC detected using the spectrophotometric method. On the other hand, the TRAP method measures the reaction of total antioxidants and peroxyl radicals generated by 2,2'-azobis(2-amidinopropane hydrochloride; ABAP) 27. The contribution of albumin to the TRAP reaction is very low (<10%), whereas that of UA is higher (47–57%, but even up to 76.4%). Therefore, change in serum UA may influence serum TAC levels determined using the TRAP method.

Our findings also indicate a significant increase in serum UA and a positive correlation between serum UA and TAC in patients with severe sepsis and septic shock. In addition, serum UA levels correlated significantly with APACHE II scores. Similarly, a study of neonatal sepsis 28 found that serum UA concentrations were increased by 51% in babies with septic shock as compared with control babies. Jabs and coworkers 29 also found that plasma UA levels increased in relation to higher APACHE II scores in critically ill patients. All of these findings suggest that serum UA is an important contributor to serum TAC, which in turn may reflect the clinical severity of severe sepsis. The mechanism of increased UA in patients with severe sepsis and septic shock is unknown. Both increased production and decreased excretion of UA may result in elevated serum level. Severe sepsis and septic shock may induce ischemia or hypoxia in multiple organs, which further increases the change in xanthine/hypoxanthine to UA through activation of xanthine oxidase in microvascular endothelium 3031. On the other hand, renal dysfunction induced by septic shock may reduce the secretion of UA from the kidneys, which may increase serum UA and further elevate serum TAC 513. In the present study, we excluded patients with previous renal dysfunction and patients undergoing hemodialysis. Therefore, our finding of increased serum UA or TAC in patients with severe sepsis or septic shock could not have been a consequence of renal failure. Whether hyperuricemia is a risk factor for severe sepsis is unknown. Our findings simply demonstrate that serum UA levels were correlated with APACHE II scores, and indicate that hyperuricemia might be associated with poorer clinical outcomes in sepsis. In this study elevation of serum UA could not represent TAC completely. More studies are needed to investigate the mechanism underlying the relation between serum UA and sepsis.

Serum TAC levels were elevated in severe sepsis, especially in nonsurvivors, and were positively correlated with clinical severity. Serum UA, which contributes largely to TAC level, plays an important role in elevating serum TAC in patients with severe sepsis. Further studies are needed to confirm our observations and elucidate the underlying mechanisms.

• Overall, serum total antioxidant capacity levels were higher in patients with severe sepsis, especially in nonsurvivors.

• Serum total antioxidant capacity could reflect clinical severity.

• Furthermore, serum uric acid levels contributed largely to high total antioxidant capacity levels.

APACHE = Acute Physiology and Chronic Health Evaluation; CI = confidence interval; ICU = intensive care unit; SIRS = systemic inflammatory response syndrome; TAC = total antioxidant capacity; TRAP = total radical-trapping antioxidant parameter; UA = uric acid.

The authors declare that they have no competing interests.

CCC conceived the study and drafted the manuscript. CCC, LIH, and MFC all contributed to the study design. CCC obtained research funding. SCS and MFC supervised the conduct of the trial and data collection. CHC and YFT recruited patients, oversaw bacteriology, and managed the data. SCS and CCS analyzed the data, including quality control. Professor Shin-Tair Wang provided statistical advice on study design. All authors contributed substantially to its revision and approved the final manuscript. MFC takes responsibility for the paper as a whole.