Stratified assessment of the role of inhaled hypertonic saline in reducing cystic fibrosis pulmonary exacerbations: a retrospective analysis.

Objective Limited data exist concerning the role of inhaled hypertonic saline (HS) in decreasing pulmonary exacerbations in cystic fibrosis (CF), especially as more advanced stages of CF lung disease were excluded in prior studies. Herein, the authors retrospectively determined the efficacy of inhaled HS in reducing CF pulmonary exacerbations when stratified according to the severity of CF lung disease. Stratification was based on the framework of the Pulmonary Therapeutics Committee's published gradation of obstructive lung physiology in CF, that is, mild (FEV(1) >70%), moderate (FEV(1) 40-70%) and severe (FEV(1) <40%) lung disease, respectively. Design A retrospective review of the Port CF database over a 3-year period performed at an academic CF care centre. Results 340 pulmonary exacerbations were identified; inhaled HS was being used in 99 of these cases. Univariate analysis demonstrated a significant reduction in pulmonary exacerbations only in mild obstruction (OR=0.09, CI 0.01 to 0.81, p=0.012); however, multivariate logistic regression that adjusted for confounding variables showed a reduction in pulmonary exacerbations across the entire spectrum of obstructive lung disease when using inhaled HS, that is, mild obstructive CF lung disease (OR=0.17, CI 0.05 to 0.58, p=0.004), moderate obstructive CF lung disease (OR=0.39, CI 0.16 to 0.93, p=0.034), as well as severe obstructive CF lung disease (OR=0.02, CI 0.001 to 0.45, p=0.015). Moreover, inhaled HS appeared reasonably well tolerated across all stages of lung-disease severity, and was discontinued in only 7% of cases (n=4) with severe lung disease. Conclusion In this study, inhaled HS appeared to reduce pulmonary exacerbations in CF lung disease at all stages of obstruction. This underscores the importance of therapeutic inhaled HS in CF lung disease, regardless of the severity of lung obstruction.

Introduction

Cystic fibrosis (CF) is characterised by decreased clearance of airway mucus that over time leads to progressive inflammatory loss of lung function consequent to infectious exacerbations.1 In this context, the landmark trial of Elkins and colleagues2 established that inhaled hypertonic 7% saline (HS) nebulised twice daily in CF lung disease improved the overall forced expiratory volume in 1 s (FEV1) and the forced vital capacity (FVC) by 68 ml and 82 ml, respectively, although it did not appreciably alter the rate of decline of FEV1. The effect of inhaled HS in decreasing the frequency of clinical exacerbations of CF lung disease was reported as a secondary outcome in this trial; a 56% relative reduction in exacerbations for the HS-treated group.2 Prior to this report, other trials had likewise demonstrated the short-term benefit of HS,3 4 but had not specifically addressed the effect of inhaled HS in decreasing pulmonary exacerbations in CF lung disease. We sought to better define the therapeutic role of inhaled HS in reducing the frequency of pulmonary exacerbations based on the severity of underlying CF lung disease; this is in accordance with the current stratified framework of evidence-based Cystic Fibrosis Pulmonary Guidelines.5 The original abstract of this study was presented at the 2009 annual scientific meeting of the American College of Chest Physicians.6

Methods

A retrospective assessment of the Saint Louis University (SLU) institutional data within the Port CF database registry was performed. The SLU CF care centre is a combined adult and paediatric academic accredited CF centre. The Port CF registry captures both clinical and epidemiological data from all CF patient during each individual visit to the CF care centre, and is recorded as per an established nationwide protocol. Approval was obtained from the Institutional Review Board at SLU. We initially identified a cohort of CF patients who presented over a 3-year period beginning January 2006, corresponding to the publication of the Elkins study.2 All episodes of pulmonary exacerbations necessitating either hospitalisation or treatment with home intravenous antibiotics were identified. Other recorded variables included utilisation of inhaled HS, airway-clearance methodologies, demographics, sputum culture results and spirometric indices. Groups were compared utilising χ2 and independent t testing for categorical and continuous variables, respectively. The severity of CF lung disease was further stratified into three groups based on the framework of the Pulmonary Therapeutics Committee's published gradation of obstructive lung physiology in CF,5 that is, mild (FEV1 >70% predicted), moderate (FEV1 40–70% predicted) and severe (FEV1 <40% predicted) lung disease, respectively. A univariate χ2 analysis was performed separately in all three subgroups to assess the effects of HS in reducing the frequency of pulmonary exacerbations; consequently, a p value of 0.016 was considered significant utilising a Bonferroni adjustment for multiple comparisons. Finally, a logistic regression was also performed in all subgroups to adjust for differences in subgroup characteristics such as age, gender, BMI, inhaled HS, sputum positivity for MRSA or Pseudomonas, as well as the spirometric FEV1 and FVC. A statistical analysis was performed using SPSS V.17.

Results

Overall, 340 pulmonary exacerbations were identified from a cohort of 424 patients. The average age of the entire cohort was 31±11 years, and 55% were male. Inhaled HS was being used in 99/340 cases with exacerbations (29%). Fifty patients (12%), 183 patients (45%) and 170 patients (42%) of the cohort were categorised as having mild, moderate and severe CF lung disease, respectively; 21 patients were uncategorised secondary to unclear documentation of the FEV1. Demographic and clinical variables are summarised in table 1. Using a univariate analysis, we found a significant reduction in pulmonary exacerbations in the subgroup of patients using HS with mild lung disease (OR=0.09, CI 0.01 to 0.81, p=0.012), whereas no reductions were found in the cohort of subjects with moderate (OR=1.33, CI 0.65 to 2.74, p=0.432) and severe lung disease (OR=5.62, CI 0.73 to 43.21, p=0.063). However, a subsequent multivariate analysis using logistic regression modelling demonstrated a reduction in pulmonary exacerbations when using HS at all stages of obstruction, that is, mild obstructive CF lung disease (OR=0.17, CI 0.05 to 0.58, p=0.004), moderate obstructive CF lung disease (OR=0.39, CI 0.16 to 0.93, p=0.034), as well as severe obstructive CF lung disease (OR=0.02, CI 0.001 to 0.45, p=0.015). These findings are summarised in tables 2, 3, with individual regression tables in the supplementary appendix. Additionally, inhaled HS was discontinued in only four cases (7%) with severe lung disease.

Table 1

Baseline epidemiological characteristics between hypertonic-saline-treated and non-treated groups

	Hypertonic saline (n=121)	No hypertonic saline (n=303)	p Value
Age, mean±SD	33±10	31±12	0.13
Male gender, n (%)	37 (31%)	196 (65%)	<0.001
Body mass index, mean±SD	20.8±0.7	21.9±5.7	0.001
Forced expiratory volume in 1 s (% predicted), mean±SD	50±12	46±23	0.016
Forced vital capacity (% predicted), mean±SD	73±9	57±24	<0.001
Use of recombinant human DNase, n (%)	121 (100%)	303 (100%)	NA
Sputum positivity for Pseudomonas aeruginosa, n (%)	60 (50%)	211 (70%)	<0.001
Sputum positivity for methicillin-resistant Staphylococcus aureus, n (%)	60 (50%)	97 (32%)	0.001
Airway-clearance device (chest vest or flutter valve), n (%)	80 (93%)	143 (97%)
Hospitalisation/home intravenous antibiotics, n (%)	41 (52%)/41 (48%)	88 (60%)/60 (40%)

Table 2

ORs using inhaled hypertonic saline (n=99) for pulmonary exacerbations (n=340) using a univariate analysis

Lung-disease severity	OR	CIs		p Value
		Lower	Higher
Mild (n=50)	0.09	0.01	0.812	0.012
Moderate (n=183)	1.33	0.65	2.74	0.432
Severe (n=170)	5.62	0.73	43.21	0.063

Numbers in bold indicate statistical significance.

Table 3

ORs using inhaled hypertonic saline (n=99) for pulmonary exacerbations (n=340) using a multivariate analysis

Lung-disease severity	OR	CIs		p Value
		Lower	Higher
Mild (n=50)	0.17	0.05	0.58	0.004
Moderate (n=183)	0.39	0.16	0.93	0.034
Severe (n=170)	0.02	0.001	0.452	0.015

Discussion

Exacerbations of CF lung disease account for appreciable morbidity and burden of this disease, which collectively greatly decrease physical functioning and psychosocial quality of life.7 Pulmonary exacerbations in particular significantly contribute to the overall cost of CF care, accounting for up to 47% of overall costs in one study.8 Accordingly, measures to decrease pulmonary exacerbations are important. Inhaled HS decreases the viscosity of pulmonary secretions and thereby improves the rheological properties of mucus secondary to hydration of the airway surface.9 In addition, HS osmotically induces a sustained increase in the airway surface liquid volume depth, possibly allowing the cilia to beat freely by recoupling the mucociliary mechanism.10 Inhaled HS for CF lung disease is currently assigned a Grade II recommendation in the Cystic Fibrosis Pulmonary Guidelines published in 2007,5 based on evidence from the above-cited trials.2–4 Notably, in the trial reporting the efficacy of HS in reducing clinical exacerbations,2 patients with severe CF lung disease (as classified by FEV1 <40% predicted) were excluded; moreover, only approximately one-third of the included participants were using nebulised rhDNase.

Here, we have assessed the effects of inhaled HS in reducing the number of pulmonary exacerbations across varying levels of lung disease severity, especially considering that more severe forms of CF lung disease had been previously excluded.2 Our study suggests that inhaled HS is beneficial in reducing exacerbations across all stages of CF lung disease using multivariate analyses, even though the univariate analysis only showed benefit in mild CF lung disease. These findings underscore the potential therapeutic benefit of initiating inhaled HS at any stage in the disease continuum. Our study group had a significantly higher usage of nebulised rhDNase (100%) and other mechanical airway clearance therapies (97%) compared with previous studies.2 Hence, we postulate that there may be an additional benefit of inhaled HS in patients who are already on established airway-clearance strategies. We also found no indication of any excessive discontinuation of inhaled HS due to cough or bronchospasm in more severe forms of CF lung disease, suggesting that it is reasonably well tolerated even in such cases.

Even so, the methodological limitations of our study must be acknowledged, principally its retrospective nature and lack of randomisation. Mild exacerbations could possibly go unreported; moreover, there was no uniform protocol in place to assess compliance with therapy. Furthermore, the two study groups had differences in epidemiological characteristics, for example, the lower proportion of sputum Pseudomonas positivity in HS users, thereby potentially lending bias to our interpretation. In this context, regression analyses in an attempt to adjust for these variable clinical and epidemiological characteristics were performed. However, ultimately, prospective randomised controlled studies with larger numbers of included participants are warranted to better assess the benefit of inhaled HS at varying stages of CF lung-disease severity.

Conclusion

Our study demonstrated that pulmonary exacerbations appear to be reduced in patients with CF lung disease of any severity during active use of inhaled HS. The beneficial effect of HS was additive to other airway-clearance measures such as nebulised rhDNase and mechanical clearance strategies. Moreover, inhaled HS appeared reasonably well tolerated. Recently, the effectiveness of alternative therapies such as nebulised mannitol for mucociliary clearance has been reported.11 Until these newer strategies are better established, we support the use of inhaled HS in CF lung disease, in the context of reducing pulmonary exacerbations. This is especially so given the potential to improve lung function, quality of life, CF-related costs and possibly mortality.1 7 8 12 13