Plasminogen activator inhibitor-1 in sputum and nasal lavage fluids increases in asthmatic patients during common colds.

To the Editor:

The possibility that recurrent asthma exacerbations associated with common colds promote airway remodeling is suggested by the finding of accelerated lung function decline over time in patients with asthma who have frequent exacerbations. The presumed cause of loss of lung function in asthma is airway remodeling. One of the inflammatory mediators thought to promote airway remodeling is plasminogen activator inhibitor-1 (PAI-1), which inhibits both the fibrinolytic system and the matrix metalloproteinase system. We have previously reported that PAI-1 is highly expressed in patients with fatal asthma and that elevated plasma levels of PAI-1 are associated with diminished forced vital capacity. We here report that common colds are associated with increased PAI-1 production in airways of asthmatic subjects.

Fifty-two asthmatic subjects, 9 subjects with allergic rhinitis, and 14 healthy controls were evaluated within 1 to 3 days of cold onset (visit 1), then between day 5 and 7 of cold symptoms (visit 2), and at 6 weeks or longer thereafter to assess baseline status (visit 3). At each visit, induced sputum and nasal lavage fluid (NLF) samples were collected and spirometry performed as described previously. The Virochip was used to detect viruses at visit 1. Allergy skin testing was performed at visit 3 to assess atopy. This study was approved by the Internal Review Boards of the University of California at San Francisco and Northwestern University. PAI-1 concentrations in sputum and NLF were determined by using ELISA (AssayPro, St Charles, Mo). All statistical analyses were performed with the Prism software, version 5 (GraphPad, San Diego, Calif). First, all the 3 groups were compared by using the Kruskal-Wallis test, and then 2-group comparisons were done with the Mann-Whitney U test. Serial measurements were analyzed with the Friedman rank test and paired comparisons by the Wilcoxon paired test. A P value of less than .05 was considered statistically significant.

Clinical characteristics of subjects showed expected differences (Table I ). A higher proportion of allergic rhinitis and asthmatic subjects were atopic compared with healthy controls, and asthmatic subjects had the lowest FEV1. There were no significant differences in age and sex among the groups. The proportion of respiratory virus detection was similar between asthmatic and healthy subjects (63.4% vs 71.4%). Among the detected viruses, rhinovirus was the most prevalent in the 3 subject groups. At baseline, sputum PAI-1 levels were significantly higher in asthmatic subjects than in nonasthmatic controls (median ± interquartile range, 3.6 ± 2.6 vs 2.3 ± 2.1 ng/mL; P < .02) (Fig 1 , A). In asthmatic patients, sputum PAI-1 levels increased significantly on day 5 to 7 compared with the baseline levels (P < .05; Fig 1, B), whereas they did not change significantly in nonasthmatic subjects (see Fig E1 in this article's Online Repository at www.jacionline.org). Sputum PAI-1 levels in asthmatic patients with exacerbation (FEV1 drop ≥10%, n = 4) were higher than in those without exacerbation (n = 17), although it was not statistically significant (6.6 vs 4.7 ng/mL on days 1-3, P = .9; 11.7 vs 4.8 ng/mL on days 5-7, P = .3). There was no significant difference in baseline NLF PAI-1 levels between asthmatic and nonasthmatic subjects (0.05 vs 0.08 ng/mL, P = .2). PAI-1 levels in NLF samples from asthmatic patients were significantly higher both at days 1 to 3 and at days 5 to 7 than at baseline (P < .001 and P < .01, respectively; Fig 1, C). Interestingly, asthmatic subjects had an early elevation in PAI-1 levels (days 1-3) in NLF samples, which was not observed in NLF samples from nonasthmatic subjects (see Fig E2 in this article's Online Repository at www.jacionline.org). To investigate whether rhinovirus, the most prevalent common cold virus, induces airway epithelial cells from asthmatic subjects to produce PAI-1, we obtained and cultured primary nasal epithelial cells in submerged medium from 7 asthmatic subjects, and treated them with either human rhinovirus (HRV) serotype 16 at multiplicity of infection of 1 or vehicle control for 48 hours. PAI-1 levels in the supernatants of infected cultures from asthmatic patients increased significantly compared with noninfected cultures (P < .05; Fig 1, D).

Table I

Demographic and clinical characteristics

	Asthma (n = 52)	Allergic rhinitis (n = 9)	Healthy control (n = 14)	P value
Age (y), median (IQR)	32.5 (16.3)	35.0 (24.5)	29.5 (7.0)	.184∗
Sex: female, n (%)	40 (77)	7 (78)	8 (57)	.315†
Atopy, n (%)	38 (73)	9 (100)	2 (14)	<.0001†
Race, n (%)
White	33 (63)	5 (56)	13 (93)
Black	6 (12)	1 (11)	0 (0)
Hispanic	5 (10)	1 (11)	0 (0)
Others	8 (15)	2 (22)	1 (7)
Baseline FEV1 % predicted	92.2‡	109.0	104.0	.0008*
Virus detection, n (%)
Rhinovirus only	22 (42)	3 (33)	7 (50)
Coronavirus only	6 (12)	0 (0)	3 (21)
RSV only	2 (4)	0 (0)	0 (0)
Enterovirus only	1 (2)	0 (0)	0 (0)
Influenza virus only	0 (0)	1 (11)	0 (0)
RSV + coronavirus	1 (2)	0 (0)	0 (0)
RSV + influenza virus	1 (2)	0 (0)	0 (0)
Any virus	33 (63)	4 (44)	10 (71)	.177†
Current asthma control meds, n (%)
ICS only	8 (15)
Singular only	1 (2)
ICS + LABA	6 (12)
ICS + LABA + others	4 (8)
No medications	33 (63)
Nasal steroids, n (%)	9 (17)	0 (0)

ICS, Inhaled corticosteroids; IQR, interquartile range; LABA, long-acting β-2-adrenergic receptor agonists; meds, medications; RSV, respiratory syncytial virus.

P from the Kruskal-Wallis test comparing all 3 groups.

P from the χ2 test comparing all 3 groups.

Asthma severity based on baseline FEV1 (spirometry was available in 44 patients out of 52): FEV1 is 80% or more in 86.4% (n = 38), 60% to 79% in 13.6% (n = 6), and less than 60% in 0% of asthmatic patients.

Fig 1

PAI-1 in airway secretions during a common cold. Baseline sputum PAI-1 levels were measured in asthmatic and nonasthmatic subjects (A, red circles—allergic rhinitis; green triangles—healthy controls). Both sputum (B) and nasal lavage (C) PAI-1 levels were also measured during colds in asthmatic patients. D, PAI-1 levels in supernatants of nasal epithelial cells from asthmatic patients 48 hours after human rhinovirus (HRV) infection.

Fig E1

Sputum PAI-1 levels of nonasthmatic subjects (healthy controls, green upward triangle; allergic rhinitis, red downward triangle) on days 1 to 3 and days 5 to 7 of the common cold onset were compared with those at baseline visit (Wilcoxon paired test, red lines indicate median value, P > .05).

Fig E2

Nasal lavage fluid levels of PAI-1 of nonasthmatic subjects (healthy controls, green upward triangle; allergic rhinitis, red downward triangle) on days 1 to 3 and days 5 to 7 of the common cold onset were compared with those at baseline visit (Wilcoxon paired test, P > .05).

Our results show that at baseline, sputum PAI-1 levels are significantly higher in asthmatic patients versus nonasthmatic controls. In addition, the common cold increased PAI-1 levels in upper and lower airways of asthmatic subjects but not in control subjects. Lastly, in vitro, HRV induced epithelial production of PAI-1. Our data on increased sputum PAI-1 levels at baseline in asthma are similar to previous reports. Previous studies suggest that PAI-1 may be related to airway obstruction by not only extracellular matrix deposition in the airway wall but also intraluminal fibrin deposition.7, 8 This may explain at least in part the mechanism by which frequent exacerbations may cause progressive airway obstruction in a subset of patients, and why reduction in FEV1 is associated with a history of frequent exacerbations in asthmatic patients. A similar study of asthmatic subjects with cold showed that there was a very high level of fibrinogen in induced sputum on day 4. We hypothesize that this highly elevated fibrinogen level in airways of asthmatic subjects can potentiate conversion to fibrin, which is not degraded because of the elevated local PAI-1 level, an occurrence that may lead to the airway obstruction. Although we could not find a negative correlation between sputum PAI-1 levels and lung function due to the small sample size, we found that 2 patients with very high sputum PAI-1 level on days 1 to 3 and days 5 to 7 (Fig 1, B) were among 4 patients who had significant asthma exacerbation with FEV1 drop of 10% or more. It would be interesting to conduct further studies on this observation. A recent study showed that sputum levels of PAI-1 were significantly higher in patients with a longer duration than in those with a shorter duration of asthma. Our results raise the hypothesis that repeated respiratory viral infections may lead to repeated transient increases in airway PAI-1 levels in susceptible asthmatic patients, which over several years could lead to accelerated remodeling and progressive airway obstruction.6, 8

In summary, this study demonstrates that lower airway PAI-1 levels are higher in asthmatic subjects than in healthy subjects and a common cold further increases upper and lower airway PAI-1 levels in asthmatic subjects. These results may explain the association between recurrent exacerbations and persistent lower airway obstruction. Further studies are needed to understand whether elevated PAI-1 levels lead to the accumulation of fibrin and extracellular matrix in airways of asthmatic patients.