• 
• 

Three Part Question

In [children with CF] does [inhaled mannitol] [significantly improve respiratory function]?

Clinical Scenario

As the paediatric registrar in clinic, you see a child with cystic fibrosis (CF). Her mother mentions reading about newer treatments available to help improve lung function and mucus clearance. You have heard of inhaled mannitol being used in this context and wonder how effective it is and how it compares to the established treatments.

Search Strategy

Medline via PubMed was the primary source of articles. The search was through ‘Clinical queries’ using the broad, sensitive filter and search terms: ‘inhaled mannitol AND CF’. Limits were ‘humans’. No age or language limits were used due to the small number of studies. Dates included 1966 to March 2012.

Secondary searches were performed on the Cochrane database, Clinical Evidence and SUMSearch.

A total of 29 papers were found via PubMed, five of which included the relevant population and outcome measures (table). The references of the these papers were scanned, along with the linked articles, but no further articles were found. No additional articles were found via secondary sources

Search Outcome

5 papers

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Minasian et al 2010	Children with moderate CF, aged 8–18 years N recruited=45 N randomised=28 N completed all sections=20 3 consecutive 12-week treatment blocks (mannitol, rhDNase, mannitol and rhDNase) with 2-week washout periods	Randomised open-label crossover study (level 1b−)	Primary=change in FEV1 Secondary=FEF25–75, FVC, pulmonary exacerbations, sputum microbiology, exercise tolerance, QOL, adverse events	Significant improvement in FEV1 post-treatment in the single treatment groups – mannitol=110 ml, p=0.055 – rhDNase=120 ml, p=0.03 – mannitol and rhDNase=30 ml, p=0.67 Non-significant trend to greater improvement in FEV1 in mannitol vs rhDNase group (50 ml, 2.8% difference, p=0.42) Greater improvement in FVC post-mannitol treatment, compared to rhDNase (p=0.053) No increase in incidence of Pseudomonas aeruginosa in sputum of mannitol-treated group No difference between treatment groups for any other secondary outcomes	Higher dropout rate—may reflect treatment burden of longer treatment blocks or reduced drug tolerance in children—8/28 withdrew post-randomisation Fewer patients in final study than in power calculation No attempt to blind subjects to treatment arm due to difficulties with disguising drug delivery Compared pre-treatment FEV1 with that prior to starting each new block to ensure adequate washout with no ‘carry-over’ effect Data for many non-significant findings not documented, making data difficult to analyse Wide CI in all treatment groups—marked variation in response to rhDNase or mannitol in different individuals Adverse events not reported No significant difference in adherence between the 3 treatment groups
Jaques et al 2008	Patients with mild to moderate CF, aged 8–48 years (N=27 under 18 years) N recruited=49 N randomised=39 N completed all sections=35 2 consecutive 2-week treatment blocks (mannitol, placebo) with 2-week washout periods Possible confounders of age, gender, protocol deviation, treatment adherence and concomitant use of rhDNase were not significantly different between treatment groups	Randomised double-blind placebo-controlled crossover study (level 1b)	Primary=change in FEV1 Secondary=FEF25–75, FVC, PEFR, respiratory symptoms, QOL, safety, sputum microbiology	mprovement in mean FEV1 (p<0.001) – mannitol=121 ml, 95% CI 56, 186 – placebo=2 ml, 95% CI −65, 65 Mannitol also gave significant improvements in – respiratory symptoms – FEF25–75 150 ml (15% rise, p<0.02) – FEV1/FVC ratio (2% rise, p<0.05) Less abnormal sounds on lung auscultation in mannitol group QOL not significantly different, although trend to improvement in mannitol group (greater in patients >14 years old). Patients reported less respiratory symptoms and reduced fatigue while in the mannitol group. No difference in sputum microbiology (p>0.5) No significant difference in FVC (p=0.1) or PEFR (p=0.3)	Fewer dropouts than other studies:· 6/49 had positive airway challenge to mannitol· 4/49 withdrew due to symptoms pre-randomisation (cough or nausea)· 4/39 dropped out post-randomisationLess adherence in placebo arm than treatment arm (6/35 vs 2/38) Improved QOL scores proportionally greater in those over 14 years of age—different age groups may have different responses and adverse reactions No data breakdown by age, so unable to perform separate analysis of response/tolerance in different age groups
Teper et al 2011	Patients with CF, aged 7–68 years (N=29 under 18 years). Baseline FEV1 40–90% predicted N recruited=85 N randomised=48 4 consecutive 2-week treatment blocks (40, 120, 240, 400 mg mannitol) with 1-week washout periods. First block was always 400 mg, order of subsequent doses randomised	Randomised open-label dose–response crossover study (level 1b−)	Primary=optimum mannitol dose for clinical improvement in lung function (FEV1, FVC) Secondary=FEV1/FVC, PEF, FEF25–75, sputum microbiology, sputum volume, QOL, respiratory symptoms	Increases in FEV1 and FVC were dose-dependent Mean change FEV1 (all ages) 40 mg=−1.57% 400 mg=8.75% (150 ml) 400 mg vs 40 mg, p<0.001 Aged<18 years: p=0.013, if >18 years: p=0.002 Mean change FVC (all ages) 40 mg=−0.9% 400 mg=8.14% (183 ml) 400 mg vs 40 mg, p<0.001 Aged<18 years: p=0.004, if >18 years: p=0.007 No change in sputum microbiology Most improvement in CFQ-R respiratory domain score seen after 400 mg treatment	No nebulised hypertonic saline, rhDNase or β blockers allowed, so all changes in FEV1 were due to mannitol No control group as dose trial. No blinding. Stated no difference between paediatric and adult subgroup responses, but individual data not given Baseline values at start of each block compared to initial baseline measures. 1-week washout may not have been long enough as post-400 mg block had higher baseline values, so some carry-over effects Small dropout may reflect open-label design 40 mg dose worsened FEV1 and FVC. Could reflect low dose being a mild broncho-irritant even with negative challenge
Bilton et al 2011	Patients with CF, aged over 6 years (N=105 under 18 years). Baseline FEV1 30–90% predicted N recruited=389 N randomised=324 N in trial=295· mannitol=177 (n=63 under 18)· control=118 (n=42 under 18)N finished double blind portion=198· mannitol=112· control=8626-week blinded treatment block (mannitol 400 mg or control), 26-week optional open-label extension N=200 had >60% treatment compliance and no major protocol violations (mannitol 111, control 89) 86% entered open-label portion	Randomised double-blind controlled study (level 1b)	Primary=change in FEV1 Secondary=FVC, FEF25–75, pulmonary exacerbations, ≥5% absolute/relative change in FEV1, rescue antibiotic use, QOL During open label phase long term safety was the primary focus	mprovement in FEV1: – mannitol=119 ml (6.5%) – control=26 ml (2.4%), p<0.001 Not affected by concomitant rhDNase. Significant change in FVC: – mannitol=129 ml – control=16 ml No significant difference in FEF25–75 Reduced incidence of pulmonary exacerbations in mannitol group (34% reduction) and reduced rescue antibiotic use, p=0.045 No significant improvement in QOL but trend towards improvement. No increase in treatment burden by taking an extra medication No difference in sputum microbiology N=38 withdrew due to adverse events (28 mannitol)	Patients in mannitol group had more treatment-related adverse events (40% vs 22%, p<0.05) but these did not cause significantly increased drop-out rate (p=0.077) Mannitol benefits sustained for up to 52 weeks Use of non-therapeutic mannitol dose (50 mg) as control meant trial was effectively blinded Larger drop-out/non-compliance rates but much longer study period Standard therapies continued (including 50% on rhDNase)—did not separate out analysis Did not perform separate subgroup analysis of paediatric population Patients who went from control group to mannitol during open label study showed comparable sustained improvements in FEV1 Reduction in pulmonary exacerbations corresponded with lung function improvements
Aitken et al 2012	Patients with CF, aged 6–53 years (N=139 under 18 years). Baseline FEV1 40–89% of predicted N recruited=342 N randomised=318 N in trial=305· 400 mg mannitol=184 (n=91 under 18)· 50 mg mannitol (control)=121 (n=62 under 18)N finished double blind portion=260· mannitol=153· control=10726-week blinded treatment block (mannitol 400 mg or control), 26-week optional open-label extension	Randomised double-blind controlled study (level 1b)	Primary=change in FEV1 after first 26 weeks Secondary=FVC, pulmonary exacerbations, quantity of sputum, QOL, PEFR, adverse events, sputum microbiology	Absolute increase in FEV1 (P=0.059): – 400 mg mannitol=106.5 ml – 50 mg mannitol=52.4 ml, Significant increase in % change in FEV1 from baseline (p<0.01) – 400 mg mannitol=8.22%, Fewer pulmonary exacerbations in treatment group but not significantly different Increased median sputum weight in treatment group, p<0.05 No difference in sputum microbiology	Nebulised hypertonic saline prohibited but all other standard therapies continued, including 75% on rhDNase—did not separate out analysis 20 patients withdrew from the trial due to adverse events (14 in 400 mg group) Use of non-therapeutic mannitol dose (50 mg) as control meant trial was effectively blinded Did not perform separate subgroup analysis of paediatric population

Comment(s)

Inhaled mannitol acts as a hyperosmolar expectorant, increasing tracheobronchial clearance in a range of conditions, including bronchiectasis and asthma (wills 2006). Multiple mechanisms of action are proposed, including a reduction in mucus viscoelasticity, increasing ciliary beat frequency and mucus breakdown, restoring the periciliary fluid layer and triggering the cough reflex (Wills 2007, Daviskas 2006, 2010). Five studies (Table)investigated use of inhaled mannitol in children with CF. Although using some similar methodologies (including treatment regimes, exclusion criteria and outcomes), the studies differed in their treatment periods, washout lengths and use of concurrent medications, thus making it difficult to collate the results and perform a combined analysis. The evidence, however, suggests inhaled mannitol can improve lung function and is superior to placebo, with all studies demonstrating statistically significant improvements in FEV1 (forced expiratory volume in 1 s), but it is unclear what degree of change is the minimal clinically important difference (Davies). Patients using inhaled mannitol also demonstrated a reduction in the number of pulmonary exacerbations, an important cause of morbidity and mortality in the CF population (Bilton).

Only one study investigated a purely paediatric population (Miniasin). Four studies (table) had mixed age populations, one of which stratified data by age and demonstrated no difference in response between paediatric and adult populations (Teper). Bronchoconstriction was a common side-effect, which may be particularly detrimental in a population with already reduced lung function, and between 4% (Aitken) and 35% (Teper)of participants failed the mannitol tolerance test. Tolerability may be improved by premedication with a β agonist such as salbutamol.(Teper) In those with a negative provocation challenge, the number and severity of side-effects varied widely between the studies. However, they were generally manageable and quality of life scores (notoriously difficult to assess in conditions such as this—for example, increased sputum production might score as a negative but is actually a positive (Bilton) all showed a trend towards improvement. Indeed, increased cough was frequently reported, and cited as the reason for study withdrawal (Miansin) but since it can be considered a component of the therapeutic effect, it may actually be considered beneficial. One theoretical concern has been that mannitol use would encourage the growth of pathogenic bacteria (as demonstrated in vitro (Bartholdson), although more recently research has indicated mannitol, combined with gentamicin, could help eradicate organisms (Allison). Each study assessed sputum for bacterial growth and found no significant change with mannitol treatment. This is especially important in the three longer trials (Miniasin, Bilton, Aitkin) as arguably 2 weeks may be too short for such a change to occur.

A major difference between studies was the effect of adjunctive rhDNase therapy, which is fully established as a single agent and well tolerated. Four studies used rhDNase; three allowed patients already using it to continue standard use (Jaques, Bilton, aitken) while one made it part of their study (Minasin). The patients on a prescribed combination of rhDNase and mannitol1 showed no improvement in FEV1 above baseline (in contrast to when the drugs were used separately), whereas the patients in the other studies already on rhDNase (Jacques,Bilton, Aitken) did show improvement when mannitol was added. It is difficult to explain this difference in results—it may reflect the small number of patients in the study by Minasian et al,a differential response of children or rhDNase naivety. Further research is needed to determine the cause.

All five studies show that, if tolerated, mannitol can have benefits, especially in particular population subsets1 with a dose-dependent improvement in lung function (Teper). Such improvements may be both rapid (seen within 2 weeks (Jacques ,Aitken) and prolonged (demonstrated for up to a year without tolerance developing (Bilton ,Aitken). However, the case for introducing routine mannitol therapy in the treatment of CF is not clear cut. The National Institute of Health and Clinical Excellence is currently appraising the clinical and cost effectiveness of inhaled dry powder mannitol in the treatment of CF, with a report due to be published in August 2012.At present, inhaled mannitol is licensed in Europe only for use in the adult population.In a population with an already high treatment burden, any new intervention should ideally demonstrate a clear benefit over current treatments. Mannitol does not appear to improve lung function more than rhDNase and the effect of combining treatments is not clear due to conflicting results. If introduced, the mannitol dry powder inhaler appears more convenient than nebulisers, but patients need a degree of respiratory function to generate the inspiratory flow for adequate drug delivery, which might be more difficult in young children or those with advanced stage lung disease. In summary, inhaled mannitol is a useful therapeutic option, if tolerated, and a therapeutic trial should be considered in any children not benefitting from other current therapies, such as rhDNase or hypertonic saline.

Editor Comment

CF, cystic fibrosis; CFQ-R, Cystic Fibrosis Questionnaire Revised; FEV1, forced expiratory volume in 1 s; rhDNase, recombinant human deoxyribonuclease; FEF25–75, forced expiratory flow; FVC, forced vital capacity; PEF, peak expiratory flow; PEFR, peak expiratory flow rate; QOL, quality of life.

Clinical Bottom Line

Mannitol improves FEV1 and forced vital capacity in patients with cystic fibrosis (Grade A) but may be no more effective than rhDNase (Grade D). Mannitol may reduce the frequency of infective respiratory exacerbations (Grade A) and amount of postural drainage required (Grade B). A mannitol trial would be an appropriate step in children not responding to rhDNase or hypertonic saline (Grade B).

References

1. Minasian C, Wallis C, Metcalfe C, et al. Comparison of inhaled mannitol, daily rhDNase and a combination of both in children with cystic fibrosis: a randomised trial. Thorax 2010;65:51–6.
2. Jaques A, Daviskas E, Turton JA, et al. Inhaled mannitol improves lung function in cystic fibrosis. Chest 2008;133:1388–96.
3. Teper A, Jaques A, Charlton B. Inhaled mannitol in patients with cystic fibrosis: a randomised open-label dose response trial. J Cyst Fibros 2011;10:1–8.
4. Bilton D, Robinson P, Cooper P, et al. Inhaled dry powder mannitol in cystic fibrosis: an efficacy and safety study. Eur Respir J 2011;38:1
5. Aitken ML, Bellon G, De Boeck K, et al. Long-term inhaled dry powder mannitol in cystic fibrosis: an international randomized study. Am J Respir Crit Care Med 2012;185:645–52
6. Wills P, Greenstone M. Cochrane Database Syst Rev 2006;19:CD002996. Inhaled hyperosmolar agents for bronchiectasis. Cochrane Database Syst Rev 2006;19:CD002996.
7. Wills PJ. Inhaled mannitol in cystic fibrosis. Expert Opin Investig Drugs 2007;16:1121–6. Inhaled mannitol in cystic fibrosis. Expert Opin Investig Drugs 2007;16:1121–6.
8. Daviskas E, Anderson SD, Jaques A, et al. Inhaled mannitol improves the hydration and surface properties of sputum in patients with cystic fibrosis. Chest 2010;137:861–8.
9. Daviskas E, Anderson SD. Hyperosmolar agents and clearance of mucus in the diseased airway. J Aerosol Med 2006;19:100–9.
10. Davies JC, Cunningham S, Alton EW, et al. Lung clearance index in CF: a sensitive marker of lung disease severity. Thorax 2008;63:96–7.
11. Bartholdson SJ, Brown AR, Mewburn BR, et al. Plant host and sugar alcohol induced exopolysaccharide biosynthesis in the Burkholderia cepacia complex. Microbiology 2008;154:2513–21. Microbiology 2008;154:2513–21.
12. llison KR, Brynildsen MP, Collins JJ. Metabolite-enabled eradication of bacterial persisters by aminoglycosides. Nature 2011;473:216–20.
13. National Institute for Health and Clinical Excellence Guidance, Cystic fibrosis—mannitol. http://guidance.nice.org.uk/index.jsp?action=byID&o=12224 (accessed May 2012). Cystic fibrosis—mannitol. NICE http://guidance.nice.org.uk/index.jsp?action=byID&o=12224 (accessed May 2012).
14. European Medicines Agency, Summary of Opinion. http://www.ema.europa.eu/docs/en_GB/document_library/ Oct 2011