It is recognized that eosinophilic airway inflammation is more likely respond to steroid treatment. However, in real-world practice, it is difficult to routinely assess airway inflammation using sputum induction because of technical and facility requirement. COPD (chronic obstructive pulmonary disease) is a heterogeneous disease and it remains a great challenge to identify patients who have eosinophilic airway inflammation and respond to steroid treatment well. A recent study demonstrated elevated plasma D-dimer was associated with acute inflammation and a significant predictor of pulmonary embolism in COPD exacerbated patients. D-dimer may potentially act as a marker of inflammation and a predictor of cardiovascular event in COPD patients. The investigators preliminary study demonstrated that exhaled nitric oxide (eNO) > 23.5 ppb is a good surrogate marker to predict eosinophilic airway inflammation in COPD patients who were newly diagnosed or untreated for at least 3 months. There were significant correlations among sputum eosinophils, eNO and serum total immunoglobulin E (IgE). Particularly, eNO predicted sputum eosinophilia (> 3%) in COPD at a sensitivity and specificity of 62% and 71% respectively. Herein, the investigators test the hypothesis that eNO may act as a biomarker to determine treatment option for COPD.

Eligible COPD patients (newly diagnosed or untreated for at least 3 months) will be enrolled at out-patient clinic after consenting by participants. Upon enrollment, exhaled NO (eNO) will be measured and patients will be categorized into 2 groups according to eNO levels: either high exhaled NO (greater than or equal to 23.5 ppb) or low eNO (< 23.5 ppb) group. In each group, patients will be randomized to receive either 2 inhalations of fluticasone/salmeterol 250/25 mcg/ pudd twice daily or 2 inhalations of tiotropium 2.5 mcg/inhalation for 12 weeks and followed at scheduled visits. Testing outcome measures include eNO, lung function, different count and mediators in induced sputum, and which will be tested as the following timings: before (baseline, week 0), and after treatment (week 4 and week 12). Rescue medication and drug compliance will be record.

Inclusion Criteria:

1. Male or female outpatients aged from 40 to 90 years
2. Current or ex-smoker, with smoking history ≧ 20 pack- years
3. Newly diagnosed or untreated (at least 3 months) COPD patients (forced expiratory volume in first second (FEV1)/forced vital capacity (FVC) < 70%) with post-bronchodilator FEV1 < 80 % predicted value.

Exclusion Criteria:

1. Concurrent allergic rhinitis, eczema, and asthma.
2. Clinically overt bronchiectasis, lung cancer, active tuberculosis, or other known specific pulmonary disease.
3. A chest X-ray indicating diagnosis other than COPD that might interfere with the study.
4. Major disease abnormalities are uncontrolled on therapy.
5. Alcohol or medication abuse.
6. Patients had lower respiratory tract infections or received systemic steroid in the 4 weeks prior to the commencement of study.
7. Women with childbearing potential during the period of trial.
8. Unable or unwilling to comply with all protocol