- P S Burge, consultant chest physician ()a,
- P M A Calverley, professor of respiratory medicineb,
- P W Jones, professor of respiratory medicinec,
- S Spencer, research fellowc,
- J A Anderson, senior statisticiand,
- T K Maslen, clinical project manager the ISOLDE study investigatorsd
- a Department of Respiratory Medicine, Birmingham Heartlands Hospital, Birmingham B9 5SS
- b Department of Medicine, University Hospital Aintree, Liverpool L9 7AL
- c Department of Physiological Medicine, St George's Hospital Medical School, London SW17 0RE
- d GlaxoWellcome Research and Development, Stockley Park West, Uxbridge, Middlesex UB11 1BT
- Correspondence to: P S Burge
- Accepted 7 February 2000
Objectives: To determine the effect of long term inhaled corticosteroids on lung function, exacerbations, and health status in patients with moderate to severe chronic obstructive pulmonary disease.
Design: Double blind, placebo controlled study.
Setting: Eighteen UK hospitals.
Participants: 751 men and women aged between 40 and 75 years with mean forced expiratory volume in one second (FEV1) 50% of predicted normal.
Interventions: Inhaled fluticasone propionate 500 μg twice daily from a metered dose inhaler or identical placebo.
Main outcome measures: Efficacy measures: rate of decline in FEV1 after the bronchodilator and in health status, frequency of exacerbations, respiratory withdrawals. Safety measures: morning serum cortisol concentration, incidence of adverse events.
Results: There was no significant difference in the annual rate of decline in FEV1 (P=0.16). Mean FEV1 after bronchodilator remained significantly higher throughout the study with fluticasone propionate compared with placebo (P<0.001). Median exacerbation rate was reduced by 25% from 1.32 a year on placebo to 0.99 a year on with fluticasone propionate (P=0.026). Health status deteriorated by 3.2 units a year on placebo and 2.0 units a year on fluticasone propionate (P=0.0043). Withdrawals because of respiratory disease not related to malignancy were higher in the placebo group (25% v 19%, P=0.034).
Conclusions: Fluticasone propionate 500 μg twice daily did not affect the rate of decline in FEV1 but did produce a small increase in FEV1. Patients on fluticasone propionate had fewer exacerbations and a slower decline in health status. These improvements in clinical outcomes support the use of this treatment in patients with moderate to severe chronic obstructive pulmonary disease.
Chronic obstructive pulmonary disease is a leading cause of morbidity and mortality worldwide,1 2 and its prevalence is rising.3 It occurs predominantly in tobacco smokers and is characterised by an increase in the annual rate of decline of forced expiratory volume in one second (FEV1).4 As lung function deteriorates, substantial changes in general health occur.5 Smoking cessation reduces the rate of decline in FEV1 in people with this disease,6 but no pharmacological intervention has been shown to modify the progression of disease or the associated decline in health status.
In at least 10% of patients with stable chronic obstructive pulmonary disease FEV1 will increase significantly after oral prednisolone.7 A large, retrospective, open study reported a reduction in the rate of decline of FEV1 in those taking oral corticosteroids.8 Recently, two studies over three years of inhaled budesonide 800 μg in mild to moderate chronic obstructive pulmonary disease found no effect of treatment on the rate of decline in FEV1.9 10 Clinical outcomes such as exacerbations, however, were infrequent and health status either showed no benefit of budesonide 9 or was not assessed.10
The inhaled steroids in obstructive lung disease in Europe (ISOLDE) study was designed to test the effect of inhaled fluticasone propionate 500 μg twice daily on the rate of decline of FEV1 and other relevant clinical outcomes.
Participants and methods
Eighteen UK hospitals participated. Patients were current or former smokers aged 40-75 years with non-asthmatic chronic obstructive pulmonary disease. Baseline FEV1 after bronchodilator was at least 0.8 litres but less than 85% of predicted normal, and the ratio of FEV1 to forced vital capacity was less than 70%. Previous use of inhaled and oral corticosteroids was permitted. Patients were excluded if their FEV1 response to 400 μg salbutamol exceeded 10% of predicted normal, they had a life expectancy of less than five years from concurrent diseases, or they used β blockers. Nasal and ophthalmic corticosteroids, theophyllines, and all other bronchodilators were allowed during the study.
The protocol was approved by each centre's local ethical committee and patients provided written informed consent.
Patients were recruited between 1 October 1992 and 31 March 1995. Eligible patients entered an eight week run-in period after withdrawal from any oral or inhaled corticosteroids. After clinic visits at 0, 4, and 8 weeks (visits 0, 1, and 2, respectively) patients were randomised to receive either fluticasone propionate 500 μg or an identical placebo twice daily administered from a metered dose inhaler and with a spacer device by using 10 tidal breaths after each of two actuations. We used a computer generated allocation schedule stratified by centre (block size of six). Patients were randomised sequentially from a list comprising treatment numbers only. Throughout the trial patients used salbutamol (100 μg/puff) or ipratropium bromide (40 μg/puff), or both, for symptomatic relief.
Before the double blind phase, and if not contraindicated, patients received oral prednisolone 0.6 mg/kg/day for 14 days, after which spirometry was performed. These data were used to test whether the acute corticosteroid response could predict those patients who would benefit from long term inhaled corticosteroids. During the three year double blind phase, participants visited a clinic every three months for spirometry, recording of exacerbations, and safety assessments.
The primary end point was the decline (ml/year) in FEV1 after bronchodilator. About 450 patients with two or more measurements of FEV1 during treatment were required to detect a treatment difference of 20 ml/year, assuming a linear decline and a SD of 75 ml/year, with 80% power. Other key end points were frequency of exacerbation, changes in health status, withdrawals because of respiratory disease, morning serum cortisol concentrations, and adverse events.
Spirometry measurements were recorded by well trained staff using a standardised procedure on new Sensormedics 2130D spirometers. Quality control included a computer generated check against the ATS criteria 11 and a central manual check for acceptability and reproducibility for all measurements, resulting in standards comparable with the lung health study.12 Visits were rescheduled to four weeks after any respiratory infections or exacerbations of the disease.
An exacerbation was defined as worsening of respiratory symptoms that required treatment with oral corticosteroids or antibiotics, or both, as judged by the general practitioner; specific symptom criteria were not used. Patients were withdrawn from the study if the number of exacerbations that required corticosteroids exceeded two in any three month period.
Health status was assessed at baseline and six monthly thereafter by using the disease specific St George's respiratory questionnaire (SGRQ).13 This questionnaire is sensitive to changes in treatment.14 A change in total score of four or more units represents a clinically important change in the patient's condition.5 Serum cortisol concentrations were measured before randomisation (baseline) and every six months during treatment. Samples were taken between 8 am and 10 am and were analysed with the ELISA-Boehringer Mannheim ES700 method.
At each visit patients were questioned about smoking status. Non-smoking was checked with exhaled carbon monoxide and urinary cotinine measurements. Self declared non-smokers were classified as smokers if cotinine was >40 ng/ml or carbon monoxide was >10 ppm at two visits. For analysis patients were categorised as continuous smokers, continuous former smokers, or intermittent smokers during the study.
Analyses for each parameter included all randomised patients with at least one valid measurement. To use all patient data we adopted the mixed models approach 15 for the primary analysis of FEV1 and total score. This is the most suitable technique for estimating rates of change, with allowance for the correlation structure of repeated measures data. Regression estimates were adjusted for patient differences in the number of observations contributing to the model and for variances within patients.16 Fixed effects were time and five covariates: baseline value centre, age, sex, and smoking status. Baseline FEV1 was the mean at four and eight weeks of the run-in period—that is, at least four weeks after withdrawal of corticosteroids. Subject effects were assumed to be random. The treatment by time interaction tested for a differential treatment effect on the rate of change in FEV1 or respiratory questionnaire score. The model for FEV1 also included a treatment main effect to help to account for the early non-linear treatment changes. Measurements at the end of the prednisolone trial were excluded from the model of decline in FEV1. FEV1 was also compared by using analysis of covariance after 3, 6, 12, 24, and 36 months to investigate treatment differences over time.
Patient exacerbation rates were calculated as the exacerbation number per treatment days and extrapolated-interpolated to a number per treatment year. The Wilcoxon rank sum test,17 stratified by centre,tested for treatment differences.
Fisher's exact test compared treatment withdrawals due to respiratory causes. These included any non-malignant lower respiratory diseases. Analysis of covariance compared data on log transformed serum cortisol concentration during treatment, adjusted for baseline. Tests were two sided, with a 5% significance level.
Of the 751 patients randomised, 376 received fluticasone propionate and 375 placebo (figure 1). During the double blind phase, 160 patients (43%) withdrew from the fluticasone propionate group and 195 patients (53%) from the placebo group, the commonest reason being frequent exacerbations of chronic obstructive pulmonary disease. Mean FEV1 at visit two was 160 ml lower in patients who withdrew from placebo compared with those who did not withdraw (1.30 litre v 1.46 litre); patients who withdrew from fluticasone propionate had a 40 ml higher FEV1 compared with those who did not withdraw (1.44 litre v 1.40 litre). Treatment groups were well matched at baseline (table 1).
Changes in FEV1
There was a fall in mean FEV1 after bronchodilator during the the run-in (placebo 75 ml, fluticasone propionate 65 ml) (fig 2). The effect was greater in patients who withdrew from inhaled corticosteroids at run-in (89 ml compared with 47 ml in the steroid naive group). After oral prednisolone there was a 60 ml (SD 170 ml) improvement in mean FEV1 after bronchodilator in both treatment groups. Subsequently mean FEV1 declined gradually in the fluticasone propionate group whereas in the placebo group it fell within three months to values before prednisolone treatment.
The annual rate of decline in FEV1 was 59 ml/year in the placebo group and 50 ml/year in the fluticasone propionate group (P=0.16) (table 2). This small difference in slopes was uninfluenced by smoking status, age, sex, or FEV1 response to the oral corticosteroid trial. The predicted mean FEV1 at three and 36 months in the fluticasone propionate group was 76 ml and 100 ml higher, respectively, than in the placebo group (mixed effects model P<0.001). The analysis of covariance showed that FEV1 in the fluticasone propionate group was higher than in the placebo group by at least 70 ml at each time point (P0.001). There was no significant relation between FEV1 response to oral corticosteroid or fluticasone propionate (P=0.056).
The median yearly exacerbation rate was lower in the fluticasone propionate group (0.99 per year) compared with the placebo group (1.32 per year), a reduction of 25% in those receiving fluticasone propionate (P=0.026).
At baseline the total respiratory questionnaire score was not significantly different between treatment groups (table 1), and it did not change significantly over the first six months of treatment (placebo: up 1.2 (SD 11.9); fluticasone propionate: down 0.5 (SD11.8); P=0.09). Thereafter it increased (that is, health status declined) over time (figs 3 and 4). This increase was linear (P<0.0001). The respiratory questionnaire score worsened at a faster rate (P=0.004) with placebo (3.2 units/year) than with fluticasone propionate (2.0 units/year).
More patients in the placebo group than in the fluticasone propionate group withdrew because of respiratory disease that was not associated with malignancy (25% v 19%, respectively; P=0.034).
Reported events were similar between treatments (table 3), except for a slightly higher incidence of events related to inhaled glucocorticoid in the fluticasone propionate group.
There was a significant (P0.032) yet small decrease in mean cortisol concentrations with fluticasone propionate compared with placebo (table 4). No more than 5% of patients on fluticasone propionate had values below the normal range during the study at any time. No decreases were associated with any signs or symptoms of hypoadrenalism or other clinical effects.
Inhaled corticosteroids have been used widely in the United Kingdom for the empirical treatment of symptomatic chronic obstructive pulmonary disease, but evidence to support this practice is limited. Unlike early reports,18 19 our study in moderate to severe chronic obstructive pulmonary disease found no effect of corticosteroids on the rate of decline in FEV1—a finding consistent with two recent budesonide studies in mild disease.9 10 Like Euroscop, a study in continued smokers,10 we found a small improvement in FEV1 after bronchodilator at three months, which was maintained throughout the study. The clinical significance of this change in airway function is unclear. Our study also showed no significant relation between corticosteroid trial response and response to long term inhaled corticosteroid.
The exacerbation rate for placebo was similar to that seen in previous reports,20 but for fluticasone propionate it was 25% lower. Precise definition of an exacerbation is difficult in ambulant patients with chronic obstructive pulmonary disease, but, by using the operational approach adopted in ISOLDE, reductions in exacerbation severity were seen in another study of patients with moderately severe disease treated for six months with fluticasone propionate.21 During the ISOLDE run-in we also observed that withdrawal of inhaled corticosteroids was associated with an increased likelihood of an exacerbation.22 These observations suggest that inhaled corticosteroids do modify the risk of symptomatic deterioration in chronic obstructive pulmonary disease.
Assessment of of health status is recognised as an important additional measurement in patients with chronic respiratory disease and is a better predictor of admission to hospital and death within 12 months than FEV1.23 The baseline respiratory questionnaire score showed significant impairment, in keeping with other populations with similar reductions in FEV1.13 14 This study shows for the first time that, like FEV1, health status declines at a measurable rate in patients with moderate to severe chronic obstructive pulmonary disease. Fluticasone propionate significantly reduced this rate of decline, delaying the average time for a clinically important reduction in health status from 15 to 24 months. As the respiratory questionnaire has only a weak correlation with FEV1,5 it must be reflecting other disease components other than airflow limitation.
What is already known on this topic
Inhaled corticosteroids are widely prescribed for patients with chronic obstructive pulmonary disease, although there are few studies to support this.
A meta-analysis of three small studies showed improvements in FEV1 with high dose beclomethasone dipropionate or budesonide but no benefit from medium dose treatment.
In two recent large studies, budesonide in medium dose produced either no benefit or a small initial improvement in FEV1
What this study adds
This study measured progressive decline in health status of patients with chronic obstructive pulmonary disease rather than just the FEV1
In patients with moderate to severe disease, fluticasone propionate 1 mg daily resulted in fewer exacerbations, a reduced rate of decline in health status, and higher FEV1 values than placebo treatment.
Serious side effects were similar to placebo, topical side effects were increased.
These data provide a rationale for the use of high dose inhaled corticosteroids in patients with moderate to severe chronic obstructive pulmonary disease.
Several factors, including disease severity, comorbidity, and study duration, contributed to the high withdrawal rate. Patients were also actively withdrawn from the study and not subsequently followed up if they experienced frequent exacerbations; this is an acknowledged limitation of the study. The effect of the differential rate of withdrawal from treatment is difficult to quantify, nevertheless it is likely to have led to a conservative estimate of benefit with fluticasone propionate.
Reports of adverse events for each treatment were generally similar, although the incidence of events related to glucocorticoids was slightly higher in the fluticasone propionate group. The incidence of fractures was low (2%) and similar to that reported in Euroscop.10 No more than 5% of patients on fluticasone propionate had cortisol concentrations below the normal range at any time during treatment. Similar reassuring data have been reported from a two year placebo controlled study of fluticasone propionate 500 μg twice daily in adults with mild asthma.24
We found no benefit of fluticasone propionate on the rate of decline in FEV1, although small improvements in FEV1 were seen. Unlike the two studies in patients with milder disease, where other clinical outcomes were less measurable,9 10 we found that fluticasone propionate 500 μg twice daily significantly reduced exacerbations and the rate of decline in health status. These data provide a rationale for the current practice of using use of inhaled corticosteroids at this dose in patients with moderate to severe chronic obstructive pulmonary disease.
Dr G F A Benfield, Dr M D L Morgan, Dr J C Pounsford, Dr R M Rudd, and Professor S G Spiro provided input into the design of the study. The scientific committee members comprised: Dr G F A Benfield, Professor P M A Calverley, Dr J Daniels, Dr A Greening, Professor G J Gibson, Professor P W Jones, Dr M D L Morgan, Dr R Prescott, Dr J C Pounsford, Dr R M Rudd, Professor D Shale, Professor S G Spiro, Mrs J Waterhouse, Dr J A Wedzicha, and Dr D Weir. The steering committee members were Mrs G Bale, Dr P S Burge, Professor P W Jones, and Dr G F A Benfield. Quality control of spirometry data was supervised by Jonathon Daniels and Geraldine Bale, who also acted as study nurse coordinator. Contributions in recruiting patients and with data collection were provided by Professor J G Ayres, Mrs G Bale, Dr N Barnes, Mrs C Baveystock, Dr G F A Benfield, Ms K Bentley, Dr Birenacki, Ms G Boar, Dr P Bright, Ms M Campbell, Ms P Carpenter, Ms S Cattell, Dr I I Coutts, Dr L Davies, Ms C Dawe, Ms J Dowselt, Ms K Dwyer, Mrs C Evans, Ms N Fasey, Dr A G Fennerty, Dr D Fishwick, Ms H Francis, Dr T Frank, Mrs D Frost, Professor G J Gibson, Dr J Hadcroft, Dr M G Halpin, Mrs O Harvey, Dr P Howard, Dr N A Jarad, Ms J Jones, Dr K Lewis, Mrs F Marsh, Mrs N Martin, Dr M D L Morgan, Ms L Morgan, Mrs W McDonald, Ms T Melody, Dr R D H Monie, Dr M F Muers, Dr R Niven, Dr C O'Brien, Ms V O'Dwyer, Ms S Parker, Dr M Peake, Dr W H Perks, Professor C A C Pickering, Dr J C Pounsford, Mrs K Pye, Mr G Rees, Ms A Reid, Ms K Roberts, Mrs C Robertson, Dr R M Rudd, Ms S Rudkin, Mr S Scholey, Dr P Scott, Dr T Seemungal, Ms S Shaldon, Dr C D Sheldon, Ms T Small, Professor S G Spiro, Dr J R Stradling, Ms H Talbot, Mrs J Waterhouse, Mrs L Webber, Dr J A Wedzicha, and Ms M J Wild.
Contributors: PSB and PMAC had the original idea for the present study, helped with the study design, recruited large numbers of patients, advised on data analysis, and helped with the writing of the paper. PSB chaired the scientific committee responsible for coordinating analyses, publications, and substudies. He is also the guarantor of the paper. PMAC chaired the steering committee that facilitated and monitored study progress. PWJ advised on collection and analyses of health status questionnaire data, recruited patients into the study, and helped with the writing of the paper. SS advised on data collection and carried out the health status analyses. JAA analysed the clinical efficacy data. TKM managed data collection and helped with data interpretation and the writing of the paper.
Funding GlaxoWellcome Research and Development.
Competing interests PSB has received financial support for research and attending meetings and has received fees for speaking and consulting. He also has shares in GlaxoWellcome. PMAC has received grant support and has spoken at several meetings financially supported by GlaxoWellcome. PWJ has received funds for research and members of staff from GlaxoWellcome. SS has received funds for research and members of staff from GlaxoWellcome. JAA and TKM are both employed by GlaxoWellcome. Fluticasone propionate is manufactured by Allen and Hanburys, which is owned by GlaxoWellcome.