METHODS

Section:

ChooseTop of pageAbstractMETHODS <<RESULTSDISCUSSIONReferencesCITING ARTICLES

Mechanically ventilated patients at the Regional Intensive Care Unit in the Royal Victoria Hospital, Belfast, Northern Ireland were eligible for inclusion into the study within 48 hours of the diagnosis of ALI and ARDS. ALI and ARDS were defined according to the consensus conference definition (17). Patients were screened daily, and a screening was log kept of the excluded patients. Exclusion criteria included creatine kinase (CK) > 10 times upper limit normal range, liver transaminases > 3 times upper limit normal range, patients with severe renal impairment (calculated creatinine clearance < 30 ml/min) not receiving renal replacement therapy, patients with severe liver disease (Child's Pugh score > 11), known lactose intolerance, current treatment with any lipid lowering agent including statins, contraindication to enteral drug administration; age < 18 years; pregnancy; participation in a clinical trial with an investigational medicinal product within 30 days, unlikely to survive beyond 48 hours, and declined consent. The study was approved by the local institution and the local research ethics committee, and written informed consent was obtained from the legal representative of the patient. Retrospective informed consent was obtained from the patient if possible. The Northern Ireland Clinical Research Support Centre provided clinical trials unit support for the study.

This was a single-center, prospective, double–blind, randomized, placebo-controlled clinical trial. Block randomization to 80 mg simvastatin or placebo (1:1), stratified for sepsis, was performed by an independent clinical trials statistician. The block size was unknown to the investigators. An independent clinical trials pharmacist performed treatment randomization. Blinding of the simvastatin and placebo tablets was achieved by encapsulation. The study medication was administered daily for up to 14 days if the CK was less than 10 times the upper reference range of normal and the transaminases were less than three times the upper level of normal. Treatment was continued until death, discontinuation of mechanical ventilation, patient or relative request for withdrawal of patient from the study, discontinuation of active treatment, or on Day 14.

Baseline patient demographics were recorded. Severity of illness scores were collected at baseline. To assess the etiology and severity of baseline lung injury, the Murray lung injury score (LIS) and the gas exchange, organ failure, cause, and associated conditions stratification system score were collected. The acute physiology and chronic health evaluation II (APACHE II), the simplified acute physiology score II score, and the sequential organ failure assessment (SOFA) score were recorded as global measures of disease severity.

The primary outcome measure was a reduction in extravascular lung water indexed to actual body weight (EVLWI). Extravascular lung water was also indexed to predicted body weight. EVLWI was measured daily using the single indicator transpulmonary thermodilution technique (PiCCO Pulsion Medical Systems, Munich, Germany) using a femoral arterial catheter as previously described (18).The femoral arterial catheter was left in situ for a minimum of 7 days and was removed thereafter at the discretion of the clinical team. Secondary outcome measures included oxygenation index (mean airway pressure × Fi_O2 × [100/Pa_O2]), plateau pressure, SOFA score, and the incidence of suspected unexpected serious adverse reactions. Additional ventilation parameters, hemodynamic variables, and need for renal replacement therapy were recorded. All measurements were undertaken between 9 and 11 am daily. CRP was measured three times per week. Patients were assessed daily for the occurrence of adverse events. As part of safety monitoring renal function, liver transaminases and CK were measured daily while the patient remained in the study. The incidence of ventilator-associated pneumonia was collected. Clinical outcome measures (including duration of mechanical ventilation, ICU, and hospital length of stay) and ICU and hospital survival were recorded.

When possible, plasma sampling and bronchoalveolar lavage (BAL), while the patient remained intubated, were performed at baseline, before study drug administration, and on Days 3 and 7. BAL and processing of BAL fluid and plasma were performed as described in the online supplement. Cytokines and chemokines (IL-1ra, IL-1β, TNF-α, IL-6, IL-8, IL-10, and monocyte chemoattractant protein 1) were measured using a cytometric bead array (R&D Systems, Abingdon, UK) as previously described (11). Data are presented as absolute concentrations.

Based on previous data (19), the study was powered to detect a difference of 2.2 ml/kg in EVLWI between patient treatment and control groups with 80% power at a significance level of 0.05. Data were analyzed on an intention-to-treat basis using SPSS for Windows 17 (SPSS, Inc., Chicago, IL). Data were tested for normality and are expressed as mean (SD) unless otherwise stated. Continuous data were analyzed by an unpaired t test or Mann-Whitney U test. A χ² test or Fisher's exact test was used to compare proportions. A P value of < 0.05 was considered significant.

RESULTS

Section:

ChooseTop of pageAbstractMETHODSRESULTS <<DISCUSSIONReferencesCITING ARTICLES

A total of 177 patients were identified during the screening process from September 18, 2006 to March 10, 2009. Sixty of these patients fulfilled the inclusion criteria for entry into the study; 117 patients were excluded, of whom 14 patients fulfilled more than one study exclusion. Approximately 20% of patients were excluded due to current treatment with a statin. Thirty patients were randomized to receive treatment with 80 mg simvastatin, and 30 patients were randomized to receive the placebo (Figure 1).

Figure 1. Trial profile. No patients were lost to follow up. CK = creatinine kinase.

Download Figure | Download Powerpoint

[More] [Minimize]

Table 1 compares the characteristics between study and excluded patients. The study patients recruited had similar demographics, LIS, and APACHE II scores to the excluded patients, although the excluded patients had a higher SOFA score (12.1 [3.9] vs. 10.3 [3.4]; P = 0.003).

Outcomes

There were no differences in outcomes according to whether the etiology of ALI/ARDS was due to sepsis or nonsepsis.

There was no difference in the EVLWI indexed to actual body weight at Day 7 between the simvastatin-treated group and the placebo group (13.7 [7.1] vs. 13.4 [8.0]; P = 0.90) (Figure 2). There was no difference in EVLWI indexed to predicted body weight at Day 7 between the two groups (15.4 [7.9] vs. 15.7 [10.7]; P = 0.99). There were no significant differences at Day 7 in the secondary outcomes.

Figure 2. Effects of simvastatin on extravascular lung water indexed to actual body weight (EVLWI). Patients with acute lung injury/acute respiratory distress syndrome were treated with 80 mg simvastatin (gray bars) or placebo (white bars). Simvastatin had no effect on EVLWI compared with placebo. Data are mean (SD).

Download Figure | Download Powerpoint

[More] [Minimize]

Pulmonary Function

There was a nonsignificant reduction in oxygenation index by Day 14 in the simvastatin-treated group (P = 0.08) (Figure 3A). There was a nonsignificant reduction in plateau pressure in the simvastatin-treated group by Day 14 (P = 0.09) (Figure 3B). Simvastatin resulted in a nonsignificant reduction in LIS at Day 14 when compared with placebo (P = 0.12) (Figure 3C).

Figure 3. Effects of simvastatin on pulmonary and nonpulmonary organ function. Patients with acute lung injury/acute respiratory distress syndrome were treated with 80 mg simvastatin (gray bars) or placebo (white bars). (A) Simvastatin reduces oxygenation index (OI) at Day 14 but does not reach statistical significance. Data are mean (SD). (B) Simvastatin reduces plateau pressure at Day 14 but does not reach statistical significance. Data are mean (SD). (C) Simvastatin reduces lung injury score (LIS) at Day 14 but does not reach statistical significance. Data are mean (SD). (D) Simvastatin reduces sequential organ failure assessment (SOFA) score at Day 14. Data are median (interquartile range). *P < 0.05 versus placebo.

Download Figure | Download Powerpoint

[More] [Minimize]

Pulmonary and Systemic Inflammation

By Day 3, compared with baseline, simvastatin treatment reduced BAL IL-8 by 2.5-fold (Figure 4A). BAL IL-6 was also reduced by 2.9–fold, although this was not statistically significant (Figure 4B). There was no significant decrease in other BAL cytokines and chemokines or BAL polymorphonuclear leukocyte cell counts (data not shown). Plasma CRP was reduced in both treatment groups at Day 8, but this systemic antiinflammatory effect was preserved only in the simvastatin-treated group to Day 12 (Figure 4C). There was no significant decrease in plasma cytokines and chemokines (data not shown).

Figure 4. Effects of simvastatin on pulmonary and systemic inflammation. Patients with acute lung injury/acute respiratory distress syndrome were treated with 80 mg simvastatin (gray bars) or placebo (white bars). (A) Simvastatin reduces bronchoalveolar lavage IL-8 at Day 3. Data are median (interquartile range [IQR]). (B) Simvastatin reduces bronchoalveolar lavage IL-6 at Day 3 but does not reach statistical significance. Data are median (IQR). (C) Plasma C-reactive protein (CRP) decreases in both treatment groups between baseline and Day 3. Plasma CRP is only significantly reduced in the simvastatin group at Day 14. Data are median (IQR). *P < 0.05 versus baseline.

Download Figure | Download Powerpoint

[More] [Minimize]

DISCUSSION

Section:

ChooseTop of pageAbstractMETHODSRESULTSDISCUSSION <<ReferencesCITING ARTICLES

Despite therapeutic advances in ventilation strategy and fluid management (3, 20), mortality remains high among patients with ALI/ARDS. No pharmacological agent has been proven to reduce mortality. There is emerging evidence that hydroxyl-methylglutaryl coenzyme A reductase inhibitors or statins may have a potentially beneficial role in ALI/ARDS. This randomized, double-blind, placebo-controlled clinical trial found that simvastatin hastens the resolution of pulmonary and systemic organ dysfunction. This physiological improvement was accompanied by a reduction in pulmonary and systemic inflammation, suggesting that this may be, at least in part, a potential mechanism for these physiological benefits. There was a larger improvement in cardiovascular, hematologic, and renal organ function, compared with a more limited improvement in respiratory dysfunction. Mortality from ALI is commonly due to multiorgan dysfunction, which may be mediated by endothelial dysfunction. Statins are known to have multiple effects on endothelial function (6); therefore, it can be postulated that an improvement in endothelial dysfunction with statins may mediate the improvements in systemic organ dysfunction to a greater degree than respiratory dysfunction, which is likely to be mediated by alveolar epithelial dysfunction as well.

We had anticipated that simvastatin would be associated with an improvement in organ function by Day 7. Administration of simvastatin in models of sepsis and ALI is associated with an early reduction in pulmonary and systemic inflammation (10, 11). Significantly lower CRP levels can be detected in patients with unstable angina within 48 hours of a single dose of 80 mg simvastatin (21). However, sustained improvements in physiological parameters were seen in this study only after 14 days. In contrast, the antiinflammatory effects, in keeping with previous data, were seen early in the course of treatment. This implies that, although antiinflammatory effects can occur early, more prolonged treatment may be required for these effects to be translated into improvements in organ dysfunction. This has implications for the design of future clinical trials of statins in ALI/ARDS and indicates that a more prolonged duration of treatment should be considered. The reassuring safety data from this study suggests that more prolonged treatment would be expected to be tolerated in critically ill patients.

Simvastatin caused a reduction in BAL fluid IL-8. This cytokine is implicated in the development of ALI/ARDS (22). In contrast, a recent study examining the effect of pretreatment with simvastatin in a model of ALI induced by inhaled LPS in healthy subjects found a reduction in TNF-α with a trend to decreased IL-1β (11). Although both studies found that simvastatin has antiinflammatory effects in the pulmonary compartment, the differing effects may relate to the relative temporal importance of these cytokines in the early and transient LPS model compared with established ALI/ARDS. Although plasma CRP was significantly decreased in the simvastatin-treated group, there was no decrease in plasma cytokines and chemokines measured. Although the reduction in BAL IL-8 and plasma CRP suggests an antiinflammatory response, this was not supported by a reduction in other plasma and BAL cytokines, and further research is required to confirm if simvastatin has antiinflammatory effects in ALI. High plasma concentrations of statins are achieved in patients in the ICU compared with normal control subjects. A single dose of 20 mg atorvastatin administered to critically ill patients was associated with markedly high plasma atorvastatin levels (23). Similarly, elevated plasma simvastatin levels were seen in critically ill patients with sepsis compared with healthy volunteers after administration of simvastatin 40 mg (24). Despite high plasma simvastatin concentrations, this was not reported to be associated with increased toxicity, and there was no correlation between CK levels and plasma simvastatin (25).

In keeping with these data, our study has demonstrated that administration of a high dose of simvastatin is well tolerated for up to 14 days in critically ill patients. Before this study, there were limited data available on the safety of statins in a prospective randomized clinical trial in patients with ALI. Statin administration was not associated with an elevated CK, elevated liver transaminases, or an increased need for renal replacement therapy. Indeed, simvastatin may have a renal protective effect. It is recognized that statins have a protective role in reducing proteinuria and preserving the glomerular filtration rate in patients with established cardiovascular disease (26). The effect of statins on acute kidney injury in critically ill patients is uncertain, but our data support the need for further research in this area. There were no differences in adverse events or serious adverse events between the treatment groups, and there were no serious adverse events relating to the study drug. These data support the need for future research into the role of statins in critically ill patients, including those with ALI.

Only 20% of patients who were screened in this study were already on a statin, suggesting that further clinical trials of statins in patients with ALI/ARDS are feasible.

This study has several limitations. First, patients excluded from the study had a higher SOFA score than the study participants. However, the subjects recruited are representative of the overall population of patients with ALI/ARDS, and the baseline severity of illness scores compare favorably with similar ALI/ARDS trials. Second, our study design was based on the primary outcome of improvement in EVLWI by Day 7. We therefore planned to keep the PiCCO femoral arterial catheter in situ for at least 7 days. Thereafter, the decision to remove the femoral catheter was at the discretion of the ICU physician. As a result, only 16% of the patients had EVLWI measurement at Day 14. Given that there is evidence of beneficial pulmonary and systemic effects at Day 14 in the statin group, it is possible we failed to identify a reduction in EVLWI at Day 14. Finally, this study highlights the difficulties in identifying appropriate clinical outcomes for phase 2 clinical trials in ALI/ARDS trials. Although we have shown a significant improvement in physiological measures of pulmonary and systemic function, which have been shown to be independent predictors of mortality in ALI/ARDS (27–29), these measurements are limited in that, due to the effects of death and extubation, the numbers of patients for whom these measurements are available decreases over time.

Although we have demonstrated that simvastatin has an effect on pulmonary and systemic outcomes in patients with ALI/ARDS, this phase 2 study was not powered to detect a difference in duration of ventilation or mortality, and further clinical trials powered for clinical outcomes are required. The Irish Critical Care Trials Group is planning a multicenter study of statins in ALI, and the National Institutes of Health Acute Respiratory Distress Syndrome network is recruiting patients with sepsis-induced ALI into the SAILS study (Statins for Acutely Injured Lungs), both of which are powered for clinical outcome measures. Moreover, statins might have a critical role to play as adjuvant therapy for ALI in the setting of the recent H1N1 influenza pandemic, although this study did not specifically investigate patients with ALI due to H1N1 influenza.

In conclusion, this double-blind randomized, placebo–controlled, clinical trial provides preliminary data that simvastatin may have a potentially beneficial effect in ALI/ARDS, as demonstrated by improvements in systemic organ dysfunction as well as a modest and limited improvement in pulmonary function in patients who were still ventilated at 14 days, which comprised 33% of the study population. This was associated by an early reduction in pulmonary and systemic inflammation. This supports the need for a multicenter clinical trial to evaluate the effect of statins for improving clinical outcomes and mortality in ALI/ARDS.