Pulmonary cytokine composition differs in the setting of alcohol use disorders and cigarette smoking

Abstract

Alcohol use disorders (AUDs), including alcohol abuse and dependence, and cigarette smoking are widely acknowledged and common risk factors for pneumococcal pneumonia. Reasons for these associations are likely complex but may involve an imbalance in pro- and anti-inflammatory cytokines within the lung. Delineating the specific effects of alcohol, smoking, and their combination on pulmonary cytokines may help unravel mechanisms that predispose these individuals to pneumococcal pneumonia. We hypothesized that the combination of AUD and cigarette smoking would be associated with increased bronchoalveolar lavage (BAL) proinflammatory cytokines and diminished anti-inflammatory cytokines, compared with either AUDs or cigarette smoking alone. Acellular BAL fluid was obtained from 20 subjects with AUDs, who were identified using a validated questionnaire, and 19 control subjects, matched on the basis of age, sex, and smoking history. Half were current cigarette smokers; baseline pulmonary function tests and chest radiographs were normal. A positive relationship between regulated and normal T cell expressed and secreted (RANTES) with increasing severity of alcohol dependence was observed, independent of cigarette smoking (P = 0.0001). Cigarette smoking duration was associated with higher IL-1β (P = 0.0009) but lower VEGF (P = 0.0007); cigarette smoking intensity was characterized by higher IL-1β and lower VEGF and diminished IL-12 (P = 0.0004). No synergistic effects of AUDs and cigarette smoking were observed. Collectively, our work suggests that AUDs and cigarette smoking each contribute to a proinflammatory pulmonary milieu in human subjects through independent effects on BAL RANTES and IL-1β. Furthermore, cigarette smoking additionally influences BAL IL-12 and VEGF that may be relevant to the pulmonary immune response.

Keywords: interleukins; pneumococcus; pneumonia; regulated and normal T cell expressed and secreted.

Fig. 1.

Regulated and normal T cell expressed and secreted (RANTES, also known as CCL5) present in bronchoalveolar lavage (BAL) increased as the severity of alcohol misuse increased in the 39 alcohol use disorder (AUD) subjects and controls. Alcohol misuse was characterized in severity according to the Alcohol Use Disorders Identification Test (AUDIT). A: AUDIT score vs. BAL RANTES (Spearman's correlation coefficient = ρ). B: subjects and controls were stratified into 4 zones according to AUDIT scores. Controls were contained in zone 1 (abstinence) and zone 2 (low-risk drinking), whereas AUD subjects were contained in zone 3 (moderate alcohol misuse) and zone 4 (severe alcohol misuse). Kruskal-Wallis test used in analyses.

Fig. 2.

In current smokers (n = 35), significant associations were determined to be present when cigarette-smoking duration, assessed via pack-year history, was assessed in relationship with BAL IL-1β (A), and VEGF (B). Spearman's correlation coefficient = ρ.

Fig. 3.

Correlations between cigarette-smoking intensity, assessed via packs smoked per day, were observed with BAL IL-1β (A), IL-1 receptor antagonist (IL-1RA) (B), IL-12 (C), and VEGF (D). Nonsmokers are not currently smoking (n = 20); moderate smokers currently smoke <1 pack per day (n = 10); heavy smokers currently smoke ≥1 packs per day (n = 7). Boxes indicate intraquartile range of data, with medians (50th percentile); whiskers indicate 10th and 90th percentiles of data. Kruskal-Wallis testing was used to compare medians across all 3 groups (P values provided). Post hoc within-group comparisons were also performed. Both moderate and heavy smokers had significantly different median values than nonsmokers for BAL IL-1β, IL-12, and VEGF; for IL-1RA, only moderate smokers had significantly different median values than did nonsmokers (*P < 0.05 compared with nonsmokers).

Fig. 4.

The 39 subjects were stratified by AUDs and current tobacco use: (-)AUDs and (-)smoking, n = 10; (-)AUDs and (+)smoking, n = 9; (+)AUDs and (-)smoking, n = 10; (+)AUDs and (+)smoking, n = 10. Boxes indicate intraquartile range of data, with medians (50th percentile); whiskers indicate 10th and 90th percentile of data. Significant differences between median values across the 4 groups were observed in BAL IL-1β (A), RANTES (B), IL-12 (C), and VEGF (D), (P values provided, using Kruskal-Wallis testing). Subsequent post hoc comparisons between the 4 subgroups (using the Wilcoxon method) revealed that smokers, either with or without AUDs, had significantly higher IL-1β than did non-AUD, nonsmokers. For RANTES, both AUD smokers and nonsmokers had significantly higher values than did non-AUD, nonsmokers. Conversely, smokers, either with or without AUDs, had significantly lower IL-12 and VEGF compared with non-AUD, nonsmokers. For these comparisons, *P < 0.01. Among AUD subjects only, comparing smokers to nonsmokers revealed significant differences in IL-1β, IL-12, and VEGF (#P < 0.02). Finally, among smoking subjects only, comparing AUD to non-AUD subjects revealed significant differences in RANTES (+P < 0.008).