Large population-based studies suggest that patients with chronic obstructive pulmonary disease (COPD) are 2 to 3 times at risk for cardiovascular mortality, which accounts for a large proportion of the total number of deaths. How COPD increases the risk of poor cardiovascular outcomes is largely unknown. However, there is growing evidence that persistent low-grade systemic inflammation is present in COPD and that this may contribute to the pathogenesis of atherosclerosis and cardiovascular disease among COPD patients. Inflammation and more specifically, C-reactive protein (CRP), has been linked with all stages of atherosclerosis, including plaque genesis, rupture and subsequent thrombo-fibrosis of vulnerable vessels. Recently, our group has demonstrated in a relatively small study that short-term inhaled corticosteroid (ICS) therapy can repress serum CRP levels in stable COPD patients. Conversely, withdrawal of ICS leads to a marked increase in serum CRP levels. Although very promising, these data cannot be considered definitive because the study was small in size and scope (N=41 patients). Additionally, this study did not address the potential effects of combination therapy with ICS and long-acting β2 agonists (LABA). This is an important short-coming because combination therapy of ICS and LABA have been shown to produce improved clinical outcomes over ICS monotherapy and is commonly used by clinicians in the treatment of moderate to severe COPD. We hypothesize that inhaled fluticasone (Flovent®) reduces systemic inflammation and that combination therapy (Advair®) is more effective than steroids alone in reducing systemic inflammation in COPD. In this proposal, we will implement a randomized controlled trial to determine whether ICS by themselves or in combination with LABAs can:

1. reduce CRP levels in stable COPD patients and
2. reduce other pro-inflammatory cytokines, which have been linked with cardiovascular morbidity and mortality such as interleukin-6 (IL-6) and monocyte chemoattractant protein-1 (MCP-1)

What is the problem to be addressed? Patients with chronic obstructive pulmonary disease (COPD) are at increased risk of cardiovascular events. Indeed, ischemic heart disease is one of the leading causes of mortality and hospitalization among patients with mild to moderate COPD. For every 10% decrease in forced expiratory volume in one second (FEV1), cardiovascular mortality increases by ~28%, and nonfatal coronary event increases by ~20% in mild to moderate COPD. How COPD increases the risk of poor cardiovascular outcomes is largely unknown. However, there is growing evidence that persistent low-grade systemic inflammation is present in COPD and that this may contribute to the pathogenesis of atherosclerosis and cardiovascular disease among COPD patients. Circulating levels of C-reactive protein (CRP), which has been strongly linked with poor cardiovascular outcomes in the general population, has been demonstrated to be elevated in COPD. Moreover, an elevated level of CRP has been associated with myocardial injury in COPD. Reduction in the level of CRP, on the other hand, has been shown to be associated with improved outcomes in various populations. Other cytokines such as interleukin-6 (IL-6) and monocyte chemoattractant protein-1 (MCP-1), which are potent regulators of CRP, have also been associated with cardiovascular events. If this linkage between systemic inflammation and atherosclerosis holds true for COPD, then systemic inflammation and/or its markers may provide a new and very important therapeutic target for COPD management. Corticosteroids (CS) can reduce CRP and other circulating inflammatory cytokine levels in acute pro-inflammatory states. They can also down-regulate certain inflammatory cells and cytokine expression in the airways of COPD patients and attenuate airway hyperresponsiveness related to COPD. More importantly, in large clinical studies, they have been shown to reduce clinical exacerbations, improve health status and may even reduce mortality in COPD. The mechanism by which such improvement occurs is not known. Recently, our group has demonstrated in a relatively small study that short-term inhaled corticosteroid (ICS) therapy can repress serum CRP levels in stable COPD patients. Conversely, withdrawal of ICS leads to a marked increase in serum CRP levels. Although very promising, these data cannot be considered definitive because the study was small in size and scope (N=41 patients). Additionally, this study did not address the potential effects of combination therapy with ICS and long-acting β2 agonists (LABA).This is an important short-coming because combination therapy has been shown to produce improved clinical outcomes over ICS monotherapy and is commonly used by clinicians in the treatment of moderate to severe COPD. In-vitro studies suggest that steroids and LABAs may "synergistically" down-regulate inflammation in COPD. Whether this occurs in-vivo remains largely unknown and untested.

What is the proposed trial design:

This trial will be a double blind, placebo-controlled multi-center study comparing the effects of Advair, Flovent and placebo on serum CRP in COPD. All study participants will first undergo a run-in phase during which all will be treated with Flovent 500 mcg bid. This will be followed by a withdrawal phase wherein all participants will be FREE of any ICS or LABAs for 4 weeks. After the withdrawal phase, the participants will be randomly assigned (using a computer generated algorithm) to one of three arms: placebo; Flovent; or Advair Run-In Phase (4 weeks): The use of ICS, theophyllines, and leukotriene modifiers, LABA will be prohibited and subjects will be maintained on Flovent 500 mcg bid. Regular use of tiotropium and as needed use of short-acting β2 (salbutamol) and/or anti-cholinergic (Atrovent) will be allowed.

Why is this phase needed? Management of COPD is variable. Because of the controversy surrounding the use of ICS and LABAs, some patients at enrollment will be taking these medications, while others will not. This phase is to ensure uniformity of therapy (and in particular to the use of ICS in the same dose) for all study participants.

Withdrawal Phase (4 weeks): Flovent will be discontinued and participants will also not be taking any other ICS, theophyllines, LABAs or leukotriene modifiers during this period. Regular use of tiotropium and as needed use of short-acting β2 (salbutamol) and/or anti-cholinergic (Atrovent) will be allowed.

Why is this phase needed? There are two principal reasons why this phase is needed. One way of "proving" that ICS modifies serum CRP levels is to demonstrate that withdrawal of ICS increases CRP levels and their re-introduction of ICS suppresses CRP levels. The Second reason is that in the researchers' pilot study it was found that serum CRP levels were highest when patients were off ICS for 4 weeks. To achieve the necessary statistical power for this study, a reasonably high serum CRP signal is desirable at the beginning of the randomization period.

Active Treatment Phase (4 weeks): Subjects will be randomized to one of 3 arms, placebo, Advair, or Flovent. Rescue medications (anti-cholinergics and short-acting β2) will be allowed. Participants will not be taking any other inhaled corticosteroids, theophyllines, leukotriene modifiers or LABAs during this period.

Why is this phase only 4 weeks? Exacerbations and infections can elevate CRP levels by 2 to 10 fold. The rate of normalization of CRP levels after these episodes is variable; complete normalization may not take place for several weeks after the resolution of the infective or exacerbation episode. To reduce the risk that the study participants will experience clinically apparent infections or exacerbations, we have made this phase of the study relatively short (4 weeks). The short treatment period will also reduce the effects of non or suboptimal compliance of treatment medications on CRP levels. We believe that 4 weeks of therapy will be sufficient to demonstrate the suppressive effects of Flovent and Advair, given the fact that in the pilot study, an effect of Flovent on CRP after only 2 weeks of therapy was observed.

What are the proposed practical arrangements for allocating participants to trial groups? Patients will be randomized (1:2:2) to placebo, Advair, or Flovent. Patients will first be stratified based on study site to minimize the potential impact of variation in patient care across the study sites on the endpoint of interest. We have hired an external statistician (Ms. Lieling Wu) who will prepare computer-generated randomization lists blocked by study site using permuted blocks of six. The lists will be inputted into a randomization computer. When a site coordinator has identified an eligible, consented patient, he/she will contact the central co-ordinating site at St. Paul's Hospital (SPH) for a randomization number. The randomization computer will then issue a study identification number to the study coordinator and to GlaxoSmithKline (Mississauga) for delivery of the appropriate treatment package (that contains one Flovent canister and one "unknown" puffer) to the appropriate study site within two business days. The open-labeled Flovent will be used for the "run-in" phase, while the "unknown" puffer (either placebo, Flovent or Advair) will be dispensed at the start of the "active treatment phase."

What are the proposed methods for protecting against other sources of bias ? All research personnel will all be blinded to the treatment group except for the study biostatistician, who will be responsible for the randomization computer. Only he will have access to the master file, which can link patient identifiers to the randomization number. This computer will be locked away in a secure space at the James Hogg iCAPTURE Centre in SPH and will have a password protection that only he (or his designate) can access. The study medications and placebo will be packaged and delivered identically as a diskus.

What is the proposed duration of treatment period? 4 weeks of run-in; 4 weeks of withdrawal phase and 4 weeks of active treatment phase (i.e RCT).

What is the frequency and duration of followup? Participants will be seen at enrollment, after the completion of each phase of the study and with exacerbations or infections (as defined above).

What are the proposed primary and secondary outcome measures?

Primary: The difference in CRP from start of the active treatment phase to the end of the trial between the 3 groups

Secondary: measurements of MCP-1 and IL-6; St. George's Respiratory Questionnaire, SGRQ,76 scores; FEV1

How will the outcome measures be measured at follow-up?

Blood Collection: During every visit, study personnel will take two 10 ml collection of blood from participants through venipuncture (using standard techniques). Samples will be centrifuged and the serum component will be aliquoted into special tubes (provided by the coordinating site) that contain anti-proteases. They will then be shipped (Fedexed) in regular ice immediately to the Study Coordinating Center (to arrive within 24 hours of blood collection) where they will be frozen in liquid nitrogen and stored in -70ºC freezers until analysis. To avoid delays, no samples will be taken on a Friday or a day preceding holidays. A high-sensitive solid phase enzyme-linked immunosorbent assay (ELISA) to measure serum CRP will be used. The investigators have measured over 4,000 serum samples from the Lung Health Study with this technique. In comparison with nephelometry, another commonly used technique, CRP levels with high-sensitivity ELISA is excellent. The coefficient of variation for CRP in the researchers' laboratory is ~5%. IL-6 and MCP-1 will also be measured using high-sensitivity ELISA assays. The researchers' laboratory has also experience performing these assays using serums collected from COPD patients. The investigators have previously shown that the coefficient of variation for the IL-6 assay to be 4.7% (median; interquartile range, 1.8% to 11.2%) and the MCP-1 assay to be 3.2% (median; interquartile range, 1.5% to 5.9%).

Spirometry: Spirometry will be performed in accordance with guidelines from the American Thoracic Society during each visit. At the first visit, pre and post-bronchodilator measurements will be done. For follow-up visits, only pre-bronchodilator values will be measured.

Health Status Measurements: During each visit, study participants will complete the SGRQ in person. The SGRQ was chosen because it has excellent internal consistency (Cronbach's alpha coefficient ≥ 0.76), reliability (intraclass correlation coefficient of ~85% of responses measured 6 months apart), and is an independent predictor of future risk of exacerbations and mortality in COPD. Clinically relevant thresholds for SGRQ are considered to be score changes of ≥ 4.0 units.

What is the proposed sample size? In the pilot study (described above), it was found that after 2 weeks, compared with the placebo group, those assigned to fluticasone experienced a significant decrease in CRP levels from baseline, after adjustments for baseline FEV1, age, and sex of participating patients (57.1% decrease relative to placebo; p=0.042). The (geometric) mean of CRP for this cohort was 4.9 mg/L (95% CI, 3.3 to 7.1). Sample sizes of 200 participants combined in Flovent and Advair group (100 in each) and 50 in the placebo group will be needed.

Inclusion Criteria:

• All patients must have a clinical diagnosis of chronic obstructive pulmonary disease according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines.
• Patients must have a cigarette smoking history of more than 10 pack-years
• Patients must be clinically stable and at least 4 weeks from last acute exacerbation (and return to baseline level of symptoms)
• Patients must have an FEV1 of less than 80% of predicted values with FEV1 to FVC ratio of less than 0.70 (post-bronchodilator values)
• Men or women ≥ 45 years of age