Acute lung injury (ALI) is an often deadly form of respiratory failure caused by inflammation. The widespread injury to the lungs involves a build-up of fluid in the lungs that results in damage to the lungs and a decreased ability of the lungs to function properly. There is currently no specific treatment for ALI; and of the 200,000 cases reported in the US yearly, about 40% will die.

Some patients may have other conditions that could change the risk of developing ALI. For example, it has been reported that patients who have septic shock have a lower chance of developing ALI if they also have diabetes mellitus (diabetes). Septic shock is a type of infection that causes overwhelming inflammation to occur throughout the body; it increases the chances that a person will develop ALI.

The reason diabetics are less likely to develop ALI is unclear and deserves further investigation. The investigators believe that it may be explained by differences in the way certain cells that are involved in inflammation function in the lungs. The investigators plan on looking at the relationship between the wide-spread inflammation that occurs within the bodies of diabetic and non-diabetic patients with an increased chance of developing ALI [sepsis].

To study this relationship, we will identify patients who fit the requirements of the study. The investigators will obtain blood samples and lung washings from these patients. The lung washings will be obtained by injecting clean salt water into the lungs through a tube and then pulling the water back out. When the sterile water is removed, some cells from the lungs also are removed and they can then be used for further lab studies. Determining how diabetes may change a patient's chance of developing ALI may uncover useful information that may help prevent or treat ALI in all patients at risk.

Acute lung injury (ALI) is a common and lethal form of respiratory failure, occurring in 200,000 people each year and in as many as 25% of mechanically ventilated patients. Despite multiple clinical studies, there are presently no effective medical therapies for ALI, and its mortality remains unacceptably high at about 40%. Common co-morbid conditions that can alter the susceptibility to ALI are chronic alcohol abuse and diabetes mellitus (DM). The incidence and severity of ALI is increased in critically ill patients with a history of chronic alcohol abuse. With respect to DM, the group has previously reported that a history of diabetes mellitus is associated with a decreased incidence of ALI in patients with septic shock. As the biological mechanism for this association was unknown, a cross-sectional epidemiological study was conducted by broadly examining the role of diabetes in the development of respiratory failure associated with sepsis. Using national hospitalization data, it was found that diabetics with sepsis were less likely to develop any form of acute respiratory failure than non-diabetics.

ALI occurs through a common pathway of oxidant stress and inflammation, injuring the lungs and rendering them dysfunctional for gas exchange. The alveolar macrophage (AM) is an important mediator of the inflammatory response, with roles such as phagocytosis and secretion of inflammatory cytokines that are key components of lung injury and repair. While more than one mechanism may explain the previous observations of a decreased incidence of ALI in diabetics, the data in chronic alcoholics suggests that AM dysfunction may play a role in the in the development of ALI leading us to hypothesize that DM may similarly alter the pulmonary microenvironment, thereby decreasing the susceptibility to ALI. Furthermore, early preliminary data suggests that the AM activity of peroxisome proliferator-activated receptor gamma (PPARγ), a transcription factor with anti-inflammatory properties, may be increased in diabetics. In addition, evidence suggests that macrophage activation into a pro-inflammatory versus an anti-inflammatory state may be tissue specific and altered by obesity and insulin resistance (i.e. compartmentalized in the lung). However, the effect of DM on the pulmonary inflammatory response and specifically AM function is as yet unknown. PPARγ is highly expressed in AMs, and thus PPARγ ligands have direct affects on AMs to influence lung inflammation by inhibiting oxidative burst and enhancing AM phagocytic function. PPARγ and its ligands influence AM differentiation into M1 (pro-inflammatory) and M2 (anti-inflammatory) phenotypes and a recent preclinical study demonstrated effective prevention of ALI by pretreatment with a PPARγ ligand. However, the effect that DM and PPARγ ligands have on AM phenotypic differentiation, PPARγ expression, oxidative stress and AM phagocytic function has not been well investigated. To our knowledge, there is no data on AM PPARγ activity in DM, and little is known about how this disease state may alter the pulmonary inflammatory milieu. The investigators hypothesize that DM reduces the risk of ALI by modulating AM differentiation and function via PPARγ expression, oxidative stress and AM phagocytic function, which may be positively influenced by exposure to a PPARγ ligand. Examining the impact of DM and PPARγ on AM phenotype and function may identify mechanisms responsible for the decreased incidence of ALI in diabetics; improve our understanding of the pathogenesis of a common and lethal disease, and lead to novel therapies for both diabetic and non-diabetic critically ill patients

Inclusion Criteria:

Patients within the first 72 hours of development of severe sepsis or septic shock

Exclusion Criteria:

1. No informed consent
2. Pregnancy
3. Expected survival ≤ 72 hours
4. Not mechanically ventilated
5. Age < 18 years

Patients in the ICU with severe sepsis ot septic shock that are on mechanical ventilation.