INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although alcohol is the most commonly abused drug in the world and causes dysfunction in many vital organ systems including the heart, liver, and brain, the effects of alcohol abuse on the human lung are essentially unknown (1). Individuals who abuse alcohol are predisposed to several at-risk diagnoses associated with the development of the acute respiratory distress syndrome (ARDS), such as severe trauma, and the aspiration of gastric contents. However, only recently has an epidemiological association between chronic alcohol abuse and ARDS been reported. In 1993, Jurkovich and colleagues reported that a history of chronic alcohol abuse adversely affected the outcome of trauma patients (2). The risk of pulmonary complications, defined as respiratory failure requiring mechanical ventilation, was higher in those trauma patients with a prior history of chronic alcohol abuse. More recently, we reported that a history of chronic alcohol abuse significantly increased the incidence and severity of ARDS in 351 critically ill patients with an identified at-risk diagnosis for the development of ARDS (3). This observation distinguished chronic alcohol abuse as the first reported comorbid variable that alters the risk of developing ARDS, and raises questions about the effects of chronic alcohol abuse on pulmonary structure and function.

Based upon the extensive evidence that implicates the depletion of glutathione (GSH) in the pathogenesis of ethanol-mediated liver disease, we speculated that chronic alcohol abuse may increase the susceptibility to acute lung injury through an alteration in pulmonary GSH homeostasis (4). GSH is the most abundant nonprotein thiol in living organisms and is essential for a number of vital biologic functions including the synthesis of proteins and DNA, transport of amino acids, enzyme activity, and protection of cells (5). However, the lung cannot synthesize GSH de novo. Alveolar epithelial type II cells import GSH from the plasma and concentrate it both intracellularly and in the epithelial lining fluid (ELF). Under normal conditions, the alveolar type II cell maintains a GSH concentration in the ELF of greater than 400 µM, several hundred times higher than plasma and one of the highest concentrations identified for any extracellular fluid (6).

The pathophysiology of ARDS involves multiple complex pathways and mechanisms including the production of oxygen radicals by stimulated neutrophils and macrophages (7). Under conditions of oxidant stress, such as ARDS, GSH interacts with oxygen radicals and serves to diminish the deleterious effects of these reactive species. An inadequate supply of GSH in the lung and specifically in alveolar type II cells would render these cells and the lung vulnerable to oxidative damage. Pacht and colleagues have reported that patients with ARDS have diminished concentrations of GSH in their ELF (8). In addition, a greater percentage of the ELF GSH in patients with ARDS is present in the oxidized form, suggesting that one cause of the diminished overall GSH concentration in the lung may be an increased utilization of GSH (5). However, whether decreased GSH concentrations are causative or the result of ARDS is unknown.

To test the hypothesis that chronic alcohol abuse predisposes to acute lung injury via GSH depletion, we examined the effects of chronic ethanol ingestion on pulmonary GSH homeostasis in rats (8). Ethanol-fed rats had significantly lower concentrations of GSH in the lung tissue, lung lavage fluid, and the alveolar type II cells as compared with the levels in control-fed rats. In addition, the ethanol-fed rats developed more severe lung injury after exposure to endotoxin when compared with similarly treated control-fed rats. Furthermore, GSH replacement decreased endotoxin-mediated lung injury in ethanol-fed rats, indicating a causal relationship between ethanol-induced GSH depletion and acute lung injury. However, the effects of chronic alcohol abuse on pulmonary GSH concentrations in humans have never been studied. Based upon the findings in the animal model, we hypothesized that chronic alcohol abuse may alter GSH homeostasis in the human lung. In this study, we measured the GSH concentration in the ELF and the peripheral blood obtained from individuals who chronically abuse alcohol.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject Group and Protocol

Chronic alcoholic subjects (n = 13) were recruited from the drug and alcohol dependency unit at Grady Memorial Hospital and the alcohol detoxification unit at the Veterans Administration Hospital. Patients met both of the following inclusion criteria to be enrolled: (1) Short Michigan Alcohol Screening Test (SMAST) score of greater than three, and (2) the use of alcohol up to and including the week prior to the bronchoalveolar lavage (BAL). The SMAST is a 13-item questionnaire that has been shown to detect behavioral correlates of alcoholism, focusing on the consequences of problem drinking and the subject's own perception of his or her alcohol problems. It is a sensitive screening test for identifying chronic alcohol abuse by the alcohol abuse/dependency criteria in the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (10).

In an attempt to isolate the effects of chronic alcohol abuse on pulmonary GSH homeostasis, subjects were excluded from the study if they met any of the following criteria: (1) prior medical history of cardiac disease, liver dysfunction, kidney disease, diabetes mellitus, hypertension, lung disease, human immunodeficiency virus (HIV)-positive, or gastrointestinal bleeding; (2) concomitant illicit drug use; or (3) age greater than 50 yr. None of the subjects were currently taking acetaminophen.

Thirteen healthy control subjects were also recruited for this study. All of these individuals had SMAST scores of zero. This study was approved by the Emory University human institutional review board, and informed consent was obtained from all participants.

Bronchoscopy with BAL

BAL was performed as previously described (11). A flexible fiberoptic bronchoscope (Olympus Model BF-1T20D, Melville, NY) was passed transnasally and into a subsegmental bronchus of the right middle lobe in all subjects. Once wedged, 150 ml of sterile saline (three 50-ml aliquots) were injected and immediately aspirated into 50-ml suction traps under continuous low pressure suction. Ten ml of blood were obtained from a peripheral vein within 10 min of the bronchoscopy.

BAL Processing

The BAL fluid (BALF) was immediately filtered through coarse gauze and centrifuged (1,000 rpm for 10 min) to separate cellular and noncellular elements. The supernatant was acidified to a pH of 5.5 using sulfasalicylic acid and stored at 70° C in small-volume aliquots until used in later assays. An uncentrifuged portion of the BALF was used for total cell count determined by hemocytometer. A small aliquot was air dried and stained by a Wright stain. A differential cell count was performed on a minimum of 300 cells.

Measurement of BAL GSH

The plasma samples were immediately added to 100 mM of serine bromate, pH = 8.5 containing per ml (0.5 mg of sodium heparin, 1 mg of bathophenanthroline disulfonic acid, and 2 mg of iodoacetic acid) in order to stabilize the red cell membrane and prevent red cell lysis (12). The plasma was separated from the blood by centrifugation at 1,000 g for 10 min and stored at 70° C. BALF was extracted with an equal volume of perchloric acid (5% final concentration) plus an internal standard of -glutamyl-glutamate (5 µM final concentration). Iodoacetic acid was added, the pH adjusted to 9.0 ± 0.2 and the samples incubated in the dark for 20 min to obtain S-carboxymethyl derivatives of thiols. The samples were then dansylated by adding dansyl chloride and incubating the samples for 24 h in the dark. The dansylated derivatives were then separated by high-performance liquid chromatography on a 10-µm Ultrasil amino column (13) and detected by fluorescence. The amount of reduced GSH and GSH disulfides were quantitated relative to -glutamyl-glutamate by integration. The dilution of the lung ELF by saline lavage was estimated by concomitant measurements of urea in the plasma and lavage fluid, and the ELF concentration of GSH was adjusted accordingly (14). GSH levels in the plasma and the ELF are expressed as a concentration (µM).

Statistical Methods

Using a Shapiro Wilk W test, serum GSH concentrations were normally distributed (15). However the ELF GSH concentrations were not normally distributed. To be consistent and conservative in our analyses, all of the results are reported as a median value and 25% to 75% quartiles. A Wilcoxon nonparametric test was used to determine whether the GSH concentrations were different between any two groups. An alpha value of 0.05 was used for all statistical tests.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The demographic information for the 26 subjects (13 chronic alcoholics and 13 control subjects) is included in Table 1. The mean SMAST score for the chronic alcoholic group was 9.4 ± 2.7 (a score of greater than 3, on a scale from 0 to 13, is indicative of chronic alcoholism) (10). On average, these 13 individuals used alcohol 6.6 d a week for 21 yr. In addition, nine of 13 (69%) of the chronic alcoholics had a history of cigarette smoking with an average number of pack-years (packs per day times the cumulative number of smoking years) of 22.5 yr. In the control group, six of 13 (46%) of the individuals had a history of cigarette smoking with an average number of pack-years of 18.5 yr.

Ten of the 13 chronic alcoholic individuals underwent further testing with a chest radiograph, spirometry, hematology, and chemistry studies. All of the results of the chest radiographs, spirometry, hematology studies, and electrolytes were within normal limits. The mean lactate dehydrogenase (LDH) and total bilirubin levels were also in the normal range: LDH = 199 µ/L (upper range of normal = 225 µ/L) and total bilirubin = 1.1 mg/dl (upper range of normal = 1.2 mg/dl). The mean serum glutamic-oxaloacetic transaminase (SGOT) level was slightly elevated at 56 µ/L (upper range of normal = 40 µ/L). We also determined the nutritional status of these 10 alcoholics using the Nutritional Risk Index which is derived from the serum albumin concentration and any alteration in their current weight over the last 6 mo (16). The Nutritional Risk Index was greater than 100 in all of these individuals, indicative of a normal nutritional status. All of the six smokers in the control group had normal spirometry.

BAL was performed without difficulty in all subjects, and there were no complications associated with the procedure. One hundred percent of the cells from the control subjects and chronic alcoholics were viable by trypan blue exclusion. The BAL data are included in Table 2.

Analysis of GSH Concentration in the ELF

As shown in Figure 1, the total GSH concentration in the ELF was significantly lower in the chronic alcoholic group (79 µmol [48 to 118 µmol]) as compared with the normal control group (576 µmol [493 to 728 µmol]), p < 0.001. In addition, the percentage of GSH in the oxidized form (GSSG/GSH) was higher in the chronic alcoholic group (9.8% [2.2 to 14.8%]) when compared with control subjects (2.8% [0.4 to 4.0%]); p = 0.05 (Figure 2).

View larger version (10K): [in this window] [in a new window]   Figure 1.   Total GSH concentrations in the ELF of chronic alcoholics and normal control subjects expressed as median and 25 to 75% quartiles.

View larger version (10K): [in this window] [in a new window]   Figure 2.   Percentage of GSH present in the oxidized form (GSSG/GSH) in the ELF of chronic alcoholics and normal control subjects expressed as median and 25 to 75% quartiles.

We also performed a subgroup analysis of GSH concentration in the ELF with individuals stratified by their smoking status. In the smoking subgroup (n = 15), the total GSH concentration in the ELF remained significantly lower in the chronic alcoholic smoking cohort (n = 9) (104 µmol [60 to 170 µmol]) as compared with the smoking control group (n = 6) (728 µmol [645 to 786 µmol]), p = 0.02. However, GSSG/GSH was not significantly higher in the chronic alcoholic smoking group (5.3% [2.0 to 12.2%]) when compared with smoking control subjects (3.5% [3.2 to 4.7%]); p = 0.26.

In the nonsmoking subgroup (n = 11), the total GSH concentration in the ELF was dramatically lower in the chronic alcoholic nonsmoking group (n = 4) (39 µmol [30 to 71 µmol]) as compared with the control nonsmoking group (n = 7) (510 µmol [145 to 571 µmol]), p = 0.01. GSSG/GSH was higher in the chronic alcoholic nonsmoking group (14.9% [4.0 to 16.6%]) compared with the nonsmoking control subjects (0.4% [0.3 to 2.4%]) yet this did not achieve statistical significance (p = 0.1).

Analysis of Plasma GSH Concentration

The total plasma GSH concentration in the chronic alcoholics was 2.29 µmol [0.85 to 4.48 µmol]) as compared with the normal control subjects (3.86 µmol [2.79 to 5.57 µmol]), p = 0.06. GSSG/GSH was not different between the alcoholic cohort (9.2% [6.4 to 11.3%]) and the control group (4.5% [1.0 to 14.4%]) with a p value of 0.2.

We also stratified the plasma GSH concentration according to smoking status. In the smoking cohort (n = 15), the plasma GSH concentration and GSSG/GSH were not different between the alcoholic smokers (n = 9) and the smoking control subjects (n = 6), p value of 0.4 and 0.7 respectively. In the nonsmoking cohort (n = 11), the total GSH concentration in the plasma was 1.04 µmol [0.66 to 4.73 µmol] in the chronic alcoholic nonsmoking group (n = 4) as compared with 3.86 µmol [3.69 to 6.36 µmol] in the control nonsmoking group (n = 7), p = 0.1. GSSG/GSH was higher in the chronic alcoholic nonsmoking group (7.6% [5.0 to 10.6%]) compared with the nonsmoking control subjects (1.3% [0.6 to 4.1%]) yet this did not achieve statistical significance (p = 0.07).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study reports a profound decrease in the alveolar ELF concentration of GSH in chronic alcoholic individuals. These concentrations were less than 15% of those levels found in our normal control subjects. In the alcoholic patients who were nonsmokers, the GSH concentrations approached the extremely diminished levels observed in patients with documented ARDS (5). The decreased pulmonary GSH concentrations were not solely a reflection of diminished plasma GSH concentrations as the decreased levels in the ELF were out of proportion to the decreased levels in the plasma.

These results are important for the following reasons. This is one of the first reports in humans of an association between chronic alcohol abuse and altered pulmonary function. We hypothesize that the increased incidence and severity of ARDS in the chronic alcoholic patient may be in part due to decreased GSH concentrations in the lung. The results of this study could eventually lead to the development of specific therapies that may lower the incidence of ARDS in critically ill patients with a history of chronic alcohol abuse. N-acetylcysteine (NAC), a drug that augments GSH levels, has been tried in several clinical trials of patients with ARDS with modest but encouraging results (17). In one trial of 66 patients with ARDS, NAC therapy was associated with a small increase in lung compliance compared with the control group, but did not alter the rate of survival (17). Suter and colleagues reported an improvement in oxygenation and a decreased length of ventilatory support associated with NAC therapy (18). However, these studies did not stratify patients based on their history of chronic alcohol abuse. Therefore, it is possible that GSH replacement therapy may only be beneficial in this specific patient population with a preexisting GSH deficiency.

In the United States, approximately one-half of the adult population regularly consumes alcohol, and 15 to 20 million individuals are alcoholics (1). The effects of alcohol abuse on our health care system are substantial and troubling. Twenty to forty percent of patients admitted to general hospitals have alcohol-related disorders, and hospitalizations due to alcohol abuse are as common in the elderly as those due to myocardial infarctions (20). However, there are only a few previous studies which examine the effects of chronic alcohol abuse on pulmonary function. Emirgil and colleagues reported that the total lung capacity, residual volume, forced vital capacity at one second, and diffusing capacity of carbon monoxide progressively declined with an increasing history of alcohol consumption (21). However, when other studies properly accounted for several confounding variables such as cigarette smoking, the effect of alcohol consumption on pulmonary function was not statistically significant (22). Recently, several investigators have examined the effects of alcohol abuse on more subtle activities in the lung. In guinea pigs, chronic alcohol ingestion decreased both lung surfactant levels and opsonic activity in vitro (25), and the surfactant from ethanol-fed rats had decreased bactericidal activity against pneumococcus in vitro (26).

Chronic alcohol abuse alone clearly does not result in the development of acute lung injury. However, it is possible that the GSH-deficient environment created by chronic alcohol abuse increases the likelihood of developing acute lung injury in response to oxidative stress, such as observed in septic shock. The precise etiology of this decreased GSH concentration in the ELF from the chronic ingestion of alcohol is likely to be multifactorial. Chronic ethanol administration decreases GSH synthesis in the liver and diminishes the activity of both GSH peroxidase and GSH transferase in experimental animal models (27). Jewell and colleagues reported that hepatic tissue GSH levels are decreased in chronic alcoholics whether or not there is evidence of cirrhosis (28). Because the lung cannot synthesize GSH and must import and concentrate it from the plasma, the decreased concentration in the lung may partially be indicative of a globally diminished supply of GSH. However, the plasma GSH concentrations in our study were reduced by a chronic exposure to alcohol to 60% of the normal levels, whereas the epithelial lining concentration was drastically reduced to only 14% of normal. In addition, GSH concentrations are not uniformly decreased in all organ systems. Loguercio and colleagues reported that the GSH concentration in the gastric mucosa is not reduced in cirrhotic patients who have decreased plasma GSH concentrations (29).

There are some potential modifiers that may influence the interpretation of our study. Smoking has been reported to alter the characteristics of the ELF which may confound some of our results (30). Similar to the results of Cantin and colleagues, we also found that the smoking control subjects had an increased GSH concentration (by greater than 40% in our study) in the ELF when compared with nonsmoking control subjects. The exact cause of this increase in GSH concentration in smokers is presently unknown (31). In addition, malnutrition from chronic alcohol abuse may lead to alterations in the activity of enzymes involved in the GSH cycle via cofactor deficiencies (32). However, most of our chronic alcoholic individuals actually had a normal nutritional status. Furthermore, there was a slightly higher percentage of males in our alcoholic groups (92%) when compared with control subjects (71%). However, there are no reported gender differences in plasma GSH concentrations (33). Finally we used plasma GSH concentrations as an indicator of systemic GSH homeostasis. Dass and colleagues have proposed that the red blood cell has a major role in the export and transportation of hepatic and renal stores of GSH precursors and in the regulation of the bioavailability of GSH to peripheral tissues (34). Therefore red blood cell GSH concentrations may be a more accurate measure of systemic GSH homeostasis. However, alcohol abuse has been previously reported to reduce both plasma and red cell GSH concentration to a similar magnitude (35).

Ideally, the effects of chronic alcohol abuse on GSH homeostasis would be more accurately studied by direct examination of the concentration of GSH in the alveolar type II cell. Unfortunately, this determination would require an open lung biopsy or autopsy, and therefore it is not feasible. BAL is a widely used method for sampling the distal air spaces of the lung in patients that has been employed both experimentally and clinically in the evaluation of lung diseases for more than two decades. In our animal model of chronic alcohol abuse, GSH concentrations in the ELF were reduced to less than one quarter of those observed in the normal control-fed rats; however, the GSH content in freshly isolated type II cells was less than 5% of that observed in type II cells from control-fed rats. Because the alveolar type II cell is responsible for GSH homeostasis in the ELF, BAL determinations of ELF GSH concentration are likely a useful index of type II cell GSH homeostasis.

In summary, individuals with a history of chronic alcohol abuse manifest significantly diminished concentrations of GSH in the ELF. We hypothesize that this deficiency may be one of the causes of the increased incidence of ARDS in chronic alcoholics. Future clinical and basic science studies are necessary to more fully determine the clinical importance of the effects of chronic alcohol abuse on GSH homeostasis and the pulmonary response to oxidative stress.