Albumin for Fluid Resuscitation in Patients with Sepsis: What Do We Expect For?
Wang Yu,An You-zhong,Ma Peng-lin
DOI: https://doi.org/10.3760/cma.j.issn.0366-6999.20130743
IF: 6.133
2013-01-01
Chinese Medical Journal
Abstract:Without adequate fluid replacement, patients with sepsis often experienced serious hypovolemia due to fever, vomiting, or diarrhea before admission. To make things worse, relative and absolute intravascular volume deficits could be exacerbated by sepsis-induced vasodilation, increased microvascular permeability and abnormal distribution of blood flow.1 Consequently, it led to poor tissue perfusion and facilitated the development of multiple organ failure. Therefore, fluid resuscitation is crucial for initial management of severe sepsis, by which the restoration and maintenance of adequate intravascular volume contribute greatly to hemodynamic stability, and attenuate poor perfusion-caused organ injuries.2–4 However, the choice of fluid remains controversial. Albumin, as a natural colloid, has a wide range of potentially significant properties such as effectively maintaining colloid osmotic pressure, scavenging reactive oxygen species (ROS), and anti-inflammatory response as well.5 However, albumin has not been reported to be with a clinical survival benefit over crystalloid solutions for fluid resuscitation in critically ill patients in a previous large-scale (7000 cases) randomized control trial (RCT, SAFE study: saline vs. albumin for fluid resuscitation in the critically ill).6 Interestedly, a subsequent analysis of SAFE study and an update meta-analysis showed that administration of albumin decreased the risk of death in patients with severe sepsis.7 Based mainly on these evidences, albumin was newly recommended as a choice in the initial fluid resuscitation of severe sepsis by Consensus Statement of the ESICM Task Force on Colloid Volume Therapy in Critically Ill Patients (grade 2B)8 and by Surviving Sepsis Campaign: International Guidelines for Management of Severe Sepsis and Septic Shock 2012 (grade 2C).9 In spite of the low-level recommendation, it is no doubt that doctor's prescription with albumin in resuscitation of severe sepsis will be greatly encouraged in the future. Therefore, it is worth to identify and synthesize all available unconfounded evidence of the effect on improvement of outcomes in septic patients of albumin compared to crystalloids for fluid resuscitation. SAFETY PROFILE OF ALBUMIN FOR FLUID RESUSCITATION The issue regarding the safety of albumin administration for volume expansion of critically ill patients was originally proposed in a meta-analysis published in the British Medical Journal (BMJ) by Cochrane Injuries Group Albumin Reviewer in 1998, which enrolled 30 randomized controlled trials including 1419 patients.10 The results showed that the pooled mortality was increased by 6.91% in albumin treated arm. In comparison with non-albumin colloids or crystalloids for volume expansion, albumin was associated with a significant increased risk of death in either the mixed population of critically ill patients (RR=1.68, 95% CI: 1.26-2.23), or in each subgroup cases with hypovolemia (RR=1.46, 95% CI: 0.97-2.22), burns (RR=2.40, 95% CI: 1.11-5.19) and hypoalbuminemia (RR=1.69, 95% CI: 1.07-2.67). The science editor of this journal published a comment titled with “why albumin may not work” at the end of this article.11 It was noted that “because of the increased permeability of the vessels”, which was commonly resulted from inflammatory mediators' injury, “the albumin solution becomes much less effective in maintaining plasma volume than in healthy individuals who have normal vessel”. On the opposite, “the filtration of fluids, together with proteins, out into the interstitial space is greatly increased and cannot be matched by lymphatic drainage”. This alternation not only reduces the efficacy of administering albumin solutions on volume expansion, but may adversely affect the homeostasis. Twenty-two comments had followed to argue the conclusions of this meta-analysis on BMJ from 1998 to 2004. Until 2004, the results of “SAFE study” convinced that compared with saline, volume expansion with 4% albumin did not increase the 28-day mortality in critically ill patients.6 So far, the safety profile of albumin for volume expansion has been established in critically ill patients. In 2011, Cochrane Injuries Group conducted another meta-analysis of 38 RCTs including a total of 10 842 patients.12 The results indicated that albumin administration was safe in the critically ill patients with either hypovolemia or hypoalbuminemia. However, albumin did not reduce the mortality significantly in comparison with normal saline. The same conclusions were further reported by an updated meta-analysis in 2012.13 Up to now, no large-scale RCT has been published to identify the specific safety profile of albumin for fluid resuscitation in patients with sepsis. But the subgroup analysis of SAFE study and the research of “Mortality after Fluid Bolus in African Children with Severe Infection”,14 including 1218 and 2097 randomized septic patients respectively, indicated that compared with saline, albumin did not significantly change the mortality. A meta-analysis,15 which enrolled 1729 septic patients in 14 RCTs up to December 2012, demonstrated that albumin did not lead to extra deaths. The above data suggest that albumin administration for fluid resuscitation is at least as safe as normal saline in septic patients. DOSE ALBUMIN IMPROVES THE OUTCOME OF SEVERE SEPSIS? Because of the high cost, the limited availability of human albumin, and the potential for associated adverse events such as fever, rashes or infectious diseases,16 more benefits, either in correcting some pathophysiological abnormalities, or in lowing morbidity, or improving outcomes to patients with severe sepsis should be expected in administering albumin for fluid resuscitation. Obviously, the investigators of “SAFE study” also wondered whether a subgroup of patients would achieve more benefits beyond the safety as exposed to albumin treatment or not. They found that the mortality was decreased by 4.6% in albumin arm (30.7%, 185/603) compared with the percentage in saline group (35.3%, 217/615, P=0.09) in patients with severe sepsis. After excluding the 299 patients with incomplete baseline data, however, the adjusted odds ratio (OR) was 0.71 (95% CI: 0.52-0.97, P=0.03) in a multiple Logistic regression analysis of the remaining 919 enrolled patients.7 Another concurrent meta-analysis, conducted by the same investigators revealed that fluid resuscitation with albumin could significantly decrease the risk of death in septic patients (OR=0.82; 95% CI: 0.67-1.00, P=0.047).17 Thus, it was concluded that the use of albumin-containing solutions for the resuscitation of patients with sepsis was associated with lower mortality compared with saline or other fluid resuscitation regimens.7,17 However, the conclusion of fluid resuscitation with albumin improving the outcome of severe sepsis was far from definitive due to inadequate reporting of data. First, it was doubt that the subgroup data of 1218 patients with severe sepsis in “SAFE study” was over estimated in either the subsequent analysis or the meta-analysis article.7,17 Originally, there was no statistically significant difference in mortality rates between the two interventions, although it is a little lower in albumin group than in saline group (30.7% vs. 35.3%, P=0.09).6 However, adjusted by excluding 299 cases with missing baseline data, assignment to albumin was independently associated with a decreased odds ratio for death at 28 days (OR 0.71, 95% CI 0.52-0.97, P=0.03) in the subsequent analysis.7 Definitely, the statistical significance was achieved by excluding cases with rate of death higher in albumin treated group (31.8% vs. 30.7%), but lower in the saline group (34.5% vs. 35.3%) than the originals. Although the mortalities still reminded similar before and after adjusting the data in both of two groups (30.3% vs. 30.7% in albumin; 35.5% vs. 35.3% in saline), the opposite movement resulted in the statistical significance (P=0.09, P=0.03). The problem is that the remained cases were not randomly assigned into each group any more. In fact, all the excluded cases had been faithfully completed fluids resuscitation protocols in both of two groups. We think that randomization might be more potential than the missed variables to impact the real mortality. Furthermore, in the meta-analysis,17 the subgroup data of 1218 patients with severe sepsis in “SAFE study” accounted for 67.91% of weight. If the 6 studies of Boldt and 3 pediatric studies were excluded, the remaining 7 of 17 RCTs contributed only 9.52% of the weight, and with low level of quality. As shown in anther meta-analysis, fluid resuscitation with albumin did not significantly reduce the mortality in the adult patients with severe sepsis after excluding Dr. Boldt's data.15 Thus, the results of sub-group analysis should be interpreted with caution because the chance of sampling may lead to difference between sub-groups which does not exist actually.18 The second, an unpublished results of the EARSS study (Early Albumin Resuscitation during Septic Shock) conducted in 28 ICUs in French showed that only 2.2% difference in mortality among the 792 enrolled patients with septic shock was found between the two interventions of either receiving 20% albumin 20 g/q8h or receiving normal saline 100 ml/q8h within 3 days after admission to ICU (24.1% vs. 26.3%, P=0.43).19 We synthesized the data of the EARSS study and the subgroup data of “SAFE study” for further analysis. The results still indicated that albumin for fluid resuscitation did not significantly change the outcome of severe sepsis, compared with saline (Figure 1). Therefore, lack of solid data has proved the potential benefit of albumin for resuscitation to the outcome of patients with severe sepsis up to day. It is expected to solve this argument by the three large-scale RCTs in near future (i.e. EARSS; PRECISE, fluid resuscitation with 5% albumin vs. NS in early septic shock, Canadian Critical Care Trials Group; and ALBIOS, Albumin Italy Outcome Study).20,21Figure 1.: Effect of albumin fluid resuscitation on risk ratio of the 28-day mortality in patients with sepsis.WHAT DO WE EXPECT FOR ADMINISTERING ALBUMIN IN SEVERE SEPSIS? First of all, administering albumin for fluid resuscitation is valid on quick correction of serous hypoalbuminemia. Actually, hypoalbuminemia was frequently developed in septic patients. The results of EARSS study showed that the plasma level of albumin below 25 g/L was detected in 88% of the septic patients within 12 hours after admission to ICU. Further, the incidence of hypoalbuminemia (lower than 25 g/L) increased over 90% at the fourth day in saline group.19 It was also found in “SAFE study” that the daily mean plasma albumin levels were all lower than 25 g/L in the first week in the septic patients receiving saline resuscitation.6 A number of factors have been known to account for sepsis associated hypoalbuminemia, which include reduction in the capability of albumin synthesis due to hepatic dysfunction,22 albumin leakage together with the filtration of fluids out into the interstitial space resulted from the increase of capillary permeability,23,24 and dilution of aggressive initial fluid resuscitation as well.9,25 Because those factors are tightly associated with systematic inflammatory response, hypoalbuminemia is hardly to be restored under control of inflammation. As we know, however, few interventions have been proved to be clinically valid on controlling the inflammatory responses yet.26 Thus, intravenous administration of albumin is the right choice to quickly correct serious hypoalbuminemia in patients with severe sepsis today. Additionally, previous studies suggested that hypoalbuminemia was closely associated with the increase of morbidity and mortality in patients with severe sepsis.27 Mangialardi et al28 reported that the incidence of acute respiratory distress syndrome (ARDS) increased significantly in the septic cases with plasma albumin level below 20 g/L compared with the patients with normal plasma albumin level (25.0% vs. 12.3%). The mortality was significant higher in patients with low level than with normal level of albumin (41% vs. 26%).28 Chou et al29 observed the effect of daily minimum of 25 g intravenous human albumin for at least 3 days within the first week of surgery ICU stay on mortality in 133 patients with severe sepsis due to secondary peritonitis. The results showed that the survival rate was significantly improved in patients with initial plasma albumin lower than 20 g/L (P=0.002).29 The above data indicated that the use of albumin in initial fluid resuscitation, rather than crystalloids only, is more likely to provide larger benefits with regards to morbidity and mortality in septic patients with plasma albumin lower than 20 g/L, i.e. severe hypoalbuminemia. However, this concept has not been supported by any high quality RCT yet. Hopefully, publishing data of the three large-scale RCTs,19–21 mentioned above, will help to better understand the effect of albumin administration to correct severe hypoalbuminemia in the resuscitation of septic patients. Finally, administration of albumin for fluid resuscitation improves organ function somewhat in patients with severe sepsis. In “SAFE study”, better maintaining of central venous pressure (CVP) in an optimal level and quicker restoration of heart rate to normal range were observed in septic patients receiving albumin infusion.6 The results of EARSS study also demonstrated that the days of not requiring catecholamines as vasopressors were significantly increased, and Sequential Organ Failure Assessment (SOFA) scores were decreased in the albumin-treated septic shock patients within 28-day period.19 These dada indicated that sepsis induced pathophysiological alternations were likely restored, at least in part, by correction of serious hypoalbuminemia. Currently, several properties, for instance, scavenging ROS, anti-inflammation, etc, have been ascribed to albumin for fluid resuscitation in both animal and human studies.5,30 However, lack of RCT focused primarily on the impact of albumin on improvement of the injured organ function in septic patients previously. Further researches on this issue will enrich the reasonale in administering albumin for initial resuscitation in patients with severe sepsis. In conclusion, although a survival benefit over crystalloids has not been achieved, administration of albumin for fluid resuscitation has been associated with better hemodynamic stability and likely lower SOFA score in the mixed population of critically ill patients in SAFE study and EARSS trial. Interestedly, significant lower incidence of morbidity and mortality were reported in septic patients with serious hypoalbuminemia (<20 g/L) in a small-scale RCT and observational study. It is worth to further investigate the association of albumin added in initial fluid resuscitation with the improvement of the injured organ function in patients with severe sepsis.