Doxycycline, a tetracycline analogue, has shown to possess immunomodulatory properties in addition to its broad-spectrum antimicrobial activity. Doxycycline is a potent inhibitor of MMP enzymes, particularly MMP-8 and MMP-9.[6–9] Studies have demonstrated reduction of MMP-9 levels by doxycycline in lung tissues of rats.[10,11] Other immunomodulatory properties of doxycycline include suppression of neutrophil migration, antiapoptotic activity, decrease in monocyte chemoattractant protein-1, and increase in the expression/secretion of the natural inhibitor of MMP (TIMP-1).[12–14] Doxycycline has shown beneficial results in periodontitis, facial acne, recurrent oral ulceration, rheumatoid arthritis, and corneal erosions on account of its non-antimicrobial properties.[14–18] Long-term treatment with doxycycline is well tolerated. Adverse effects to doxycycline include photosensitivity and gastrointestinal disturbances.[19] It seemed reasonable to assume that drug possessing anti-inflammatory and MMP-inhibiting activity like doxycycline may have beneficial effect on lung function, systemic inflammation, and dyspnea in patients of COPD having stable symptoms. The present study was designed to study the effect of doxycycline in a dose of 100 mg once a day in patients of stable COPD.

In the present study, an interventional, randomized, observer-masked, comparative design was used to investigate the efficacy of doxycycline in addition to standard treatment in patients of stable COPD. Patients included had age >40 years, smoking history >10 pack years, and were classified as stage II or III on GOLD classification for severity. Patients presenting with acute exacerbation of COPD as defined by Anthonisen et al.,[20] and patients who had moderate/severe exacerbation in last 4 weeks as defined by severity scale proposed by Cazzola et al.,[21] were excluded. Patients with active respiratory disease other than COPD (tuberculosis, pneumonia, lung malignancies, bronchial asthma), acute systemic/local infection, cardiac and gastrointestinal disorders were excluded. Patients on systemic corticosteroids at the time of screening were also excluded. The primary endpoint was change in post-bronchodilator FEV ₁ from baseline to end of 4 weeks, and secondary endpoints were changes in forced expiratory volume (FVC), serum C-reactive protein (CRP), and medical research council (MRC) dyspnea scale after 4 weeks.

At initial screening visit, patients were assessed by history, clinical examination, pulmonary function tests, chest X-ray, and sputum examination. Data regarding smoking history and exacerbations were recorded. Patients who qualified inclusion and exclusion criteria and gave written informed consent underwent a run-in period of 4 weeks. During run-in period, all the study participants received treatment with deriphyllin (100 mg thrice a day), inhaled salbutamol and ipratropium. A run-in period of 4 weeks was selected because physiological changes associated with acute exacerbation return to baseline values in 70 to 74% patients at 4 weeks.[22] Use of inhaled corticosteroids and long-acting β₂ agonists was allowed during the run-in and study period.

After run-in period, at baseline visit, patients were assessed for values of pulmonary functions by spirometry. Assessment of dyspnea was done using MRC dyspnea scale and blood sample was collected for serum CRP estimation. Subsequently, the study participants were randomized using blocked randomization method with the help of computer software into the following two groups: Group 1 (Doxycycline group): Capsule doxycycline 100 mg once a day daily for 4 weeks in addition to run-in period treatment; Group 2 (Reference group): Continued to receive treatment received in run-in period. Systemic corticosteroids or antibiotics were not allowed during the study period until clinical necessity arouse in which the patient was then excluded from the study.

After study intervention period of 4 weeks, patients were followed up and assessed for lung functions, MRC dyspnea scale, and serum CRP. Post-bronchodilator FEV₁ and FVC were recorded after administration of 200 μg of salbutamol, using metered dose inhaler. Each subject performed a minimum three acceptable FVC measures. The greatest FEV₁ and corresponding FVC value were used in subsequent analysis. Serum CRP was estimated using immunoturbidimetry on the same day of collection of sample.

At baseline, there was a difference of 29 ml in the mean FEV₁ among two groups. At 4 weeks, mean FEV ₁ increased significantly by 153 ml in doxycycline group (P < 0.001, 95% CI = -0.190 to -0.116) and decreased by 33 ml in reference group [Table 2,Figure 2]. There was significant difference in the mean change in FEV₁ (ΔFEV₁) in both groups (P < 0.001, 95% CI = 0.140 to 0.234). There was significant increase in FEV₁ % predicted, FVC, and FVC % predicted from baseline values at 4 weeks in doxycycline group. Lung function parameters did not improve in reference group. There was significant difference in mean change in FEV₁ % predicted, FVC, and FVC % predicted among both groups. The mean percent reduction in baseline serum CRP was 45.59 ± 4.6 and 15.78 ± 5.8 in doxycycline and standard group, respectively. There was significant difference in reduction of serum CR P values between both groups. MRC dyspnea did not significantly decrease from baseline in both the groups at 4 weeks [Table 3].

Doxycycline demonstrated improvement in lung function parameters for which antimicrobial action of doxycycline is unlikely to be responsible. The present study excluded patients with acute exacerbation, as defined by Anthonisen et al.[20] This classification indicates likelihood of bacterial infection as cause of an exacerbation. The study also excluded patients who had moderate to severe exacerbations in last 4 weeks according to event-based definition given by Cazzola et al.[21] Patients having infective pathology were also excluded. A randomized controlled trial has provided evidence for significant beneficial effect of antibiotics only in those COPD patients who present with increase in all of the following cardinal symptoms: Dyspnea, increased sputum volume and sputum purulence[20] or who present with sputum purulence, and at least one of other two cardinal symptoms.[23] In conclusion, there is less possibility that antimicrobial action might have contributed to beneficial effect of doxycycline in improving lung function parameters in the present study.

Improvement in lung function parameters in present study might be the result of anti-inflammatory and MMP-inhibiting activity of doxycycline. A study demonstrated that the collagenase activity in tracheal aspirates from horses suffering from COPD was sensitive to doxycycline inhibition.[24] Study by Nordstrom et al.[14] tested clinical response to 3 months doxycycline in concert with collagenase activity in patients of rheumatoid arthritis. Significant reduction in joint score and pain visual analogue scale was seen as early as 6 weeks. In the same study, saliva samples showed significant reduction in collagenase (MMP-8) activity at 12 weeks. Doxycycline has shown improvement in pulmonary disorders in which dysregulated MMP activity is held responsible. Maugban et al.[25] studied the effect of addition of doxycycline to immunosuppressive therapy in lung transplantation patients with recurrent acute rejections or obliterative bronchiolitis (OB/BOS). This was associated with improved lung functions in serial pulmonary function tests. OB/BOS is associated with elevated MMP- 9levels and the immunomodulatory effect of doxycycline might have been responsible for the beneficial effect. Long treatment with doxycycline in patient of idiopathic pulmonary fibrosis was associated with improvement in symptoms, physiological and radiological parameters.[26] Doxycycline inhibits neutrophil collagenase (MMP-8) and MMP-9 at doses readily attainable by therapeutic doses. It achieves similar concentration in lungs and plasma.[27,28]

Airway limitation and airway inflammation are separate and independent factors in pathophysiology of COPD. CRP reflects the total systemic burden of inflammation in patients of COPD.[29] Eight-year follow-up study of large cohort with airway obstruction showed that the increased CRP levels are strong predictor of COPD hospitalization and deaths[30] Increased serum CRP levels have been associated with all-cause mortality in patients with mild to moderate COPD, reduced lung function, and greater FEV ₁ decline.[31] The present study demonstrated significant difference in reduction of baseline serum CRP levels in doxycycline and standard group. The baseline serum CRP was higher in doxycycline group, though this was not significant. Anti-inflammatory agents like corticosteroids (inhaled fluticasone) for 2 weeks reduced baseline serum CRP levels by 50% in COPD patients who had stable symptoms in previous 3 months in a study.[32] In MIDAS trial, subantimicrobial doses of doxycycline significantly reduced serum CRP level by 47% in patients with coronary artery disease.[27] The study also demonstrated significant reduction in serum IL-6. The authors concluded that reduction in CRP due to doxycycline might be due to upstream inhibition of IL-6 or direct inhibitory effect on CRP synthesis in liver or both. IL-6 is a major signaling cytokine stimulant and induces CRP production and release by liver. Similar mechanism might be responsible for reduction in CRP in the present study.

Studies have demonstrated that dyspnea scores and lung function are distinct and separate in terms of their influence on health outcome.[33] The lack of improvement in MRC dyspnea scale despite significant increase in FEV₁ might be due to following reasons. First of all, the study intervention period might not be sufficiently long enough to change the perceived respiratory disability in patients of stable COPD. Second, association of FEV₁ and MRC dyspnea scale has not been demonstrated.[34] Lastly, there could be factors other than COPD which might have contributed to dyspnea in the study participants.

Limitations of the study include short study intervention period and lack of data on effect of doxycycline on long-term clinical outcomes in COPD like symptom scores, health status, exercise tolerance, and exacerbation rates. Alternate explanation for serum CRP reduction in doxycycline group could be resolution of mild or occult infection which was not recognized prior to study. The study did not address the problem of antimicrobial resistance.