Pharmacology Journal

In Canada, chronic obstructive pulmonary disease (COPD) affects 4.3% of adults aged ≥35 years, making it the fourth most common cause of illness and death. Worldwide, COPD is also the fourth leading cause of death and the 12th leading cause of disability. Acute exacerbations are the most frequent cause of hospital admissions and death among COPD patients.

Unfortunately, recent guidelines for the diagnosis, management and prevention of COPD report that none of the existing medications for COPD has been shown to modify the long-term decline in lung function. Although the goals of COPD management include the prevention of disease progression and the reduction of mortality, they also include symptom relief, treatment and prevention of exacerbations and the improvement of health status. The step care approach proposed by the Global Initiative for Chronic Obstructive Disease (GOLD) treatment guidelines recommends theophyllines as the third line of treatment when symptoms are still persistent despite treatment with short- and long-acting inhaled bronchodilators. The recommendation for inhaled corticosteroids (ICS) use is more restricted, being recommended only for patients with frequent exacerbations, as their efficacy to improve lung function has been shown to be limited, whereas they have been shown to reduce the risk of exacerbations. In the ISOLDE trial, the annual rate of exacerbations among moderate to severe COPD patients was 1.90 (SD 2.63) for placebo group and 1.43 (SD 1.93) for patients on ICS after 3 years of treatment. In this randomized controlled trial (RCT), an exacerbation was defined as worsening of respiratory symptoms that required treatment with oral corticosteroids or antibiotics, or both.

Theophyllines have been recognized for their bronchodilating effects since the early 1950s, but have recently been shown also to have anti-inflammatory properties in patients with asthma. Several clinical trials have shown that theophyllines improve lung function and dyspnoea in patients with COPD. Theophyllines may also improve mucociliary clearance, cardiovascular function, gas exchange and exercise capacity. Despite these proven clinical benefits, the use of theophyllines in the treatment of COPD has decreased over the past years, mainly due to their narrow therapeutic range and potential for drug interactions. At the same time, the use of ICS in the treatment of COPD has increased dramatically, despite the guidelines’ recommendations and the lack of scientific evidence to demonstrate their efficacy to reduce disease progression.

As in many chronic diseases, in the treatment of COPD there is an important gap between the guidelines’ recommendations and the use of prescribed medications in clinical practice. This situation commends the evaluation of drug effectiveness in real clinical practice, since guidelines’ recommendations are mainly based on the results of RCTs that might not be easily generalized into practice. To our knowledge, there is no study comparing the effectiveness of theophyllines with other available treatment options to reduce the risk of COPD exacerbations in clinical practice. We therefore conducted a large population-based cohort study to evaluate and compare the rate of moderate to severe COPD exacerbations between users of oral theophyllines, ICS and long-acting inhaled β₂-agonists (LABA) among patients aged ≥50 years.

Source of data

This study used claims data from the health administrative database of the Régie de l’Assurance Maladie du Quebec (RAMQ) and data from the MED-ECHO database from the Canadian province of Quebec. During the study period, close to 97% of elderly (>874 000 persons aged ≥65 years) and 508 000 persons aged between 50 and 64 years were covered by the RAMQ medical services plan and the drug plan. The RAMQ Prescriptions Drugs Database contains information on prescriptions filled at community pharmacies, i.e. name, dose, form, quantity of medication dispensed, date and duration of prescription, as well as the identification and specialty of the prescribing physician. The RAMQ Medical Services Database contains claims data on medical services dispensed either at hospitals, emergency departments or medical clinics, i.e. date of service, where the service was dispensed, diagnosis coded with International Classification of Diseases (ICD)-9 codes, as well as specialty and identification of the treating physician. The RAMQ database also contains patients’ socio-demographic data, such as age, gender and date of death, as well as a variable indicating whether a patient is receiving social assistance or a supplement added to the Old Age Security pension. The MED-ECHO database contains information on all admissions to acute care hospitals in Quebec, such as date of admission, length of stay, diagnosis coded with ICD-9 codes (admission, principal and secondary), identification of hospital and treating physician. The RAMQ and MED-ECHO databases were linked using an encrypted unique patient’s identifier included in all databases. The RAMQ and MED-ECHO databases have been used extensively for epidemiological studies, and the information related to medications has been proven valid and comprehensive. Moreover, medical diagnoses related to COPD recorded in the RAMQ Medical Services database have been found to be valid for research purposes.

Study population and design

From the RAMQ databases a large cohort of COPD patients aged ≥50 years between 1 January 1995 and 31 December 1999 was selected. To be included in the cohort, patients should have: (i) filled at least six prescriptions of a short-acting inhaled β₂-agonists (SABA) (epinephrine, orciprenaline, salbutamol, terbutaline, fenoterol or pirbuterol) or of an ipratropium bromide in the year preceding cohort entry (six prescriptions per year of inhaled bronchodilator correspond to 3.6 inhalations per day, on average, which is based on the recommended minimum dose of two inhalations qid when needed); (ii) received at least one medical service for COPD (service billed for the following ICD-9 diagnostic codes: chronic bronchitis codes 491.0, 491.1, 491.2, 491.8 and 491.9; emphysema codes 492.0, 492.8; and chronic airway obstruction code 496) in the year preceding cohort entry; and (iii) been covered by the RAMQ drug plan for at least 1 year prior to cohort entry. Patients were excluded if they received oral corticosteroids as a continuous therapy (at least six filled prescriptions of ≥28 days, or at least 21 filled prescriptions regardless of duration), were hospitalized for >30 days in one hospitalization (because no information was available on medication in RAMQ database when patients were hospitalized) or received any medical service for asthma (ICD-9 codes 493) in the year preceding cohort entry. Corticosteroid-dependent patients were excluded, because this treatment might have been prescribed for a disease other than COPD and continuous use of oral corticosteroids would preclude the evaluation of our main outcome, which is based on markers of COPD exacerbations, such as a filled prescription of oral corticosteroids. Cohort entry was defined as 1 January 1996, 1997, 1998, 1999 or 2000. Patients were followed for a maximum of 7 years. Follow-up was stopped either when patients reached the end of the study period, i.e. 31 December 2002, or died. Patients were also censored if they left the RAMQ drug plan insurance, received one asthma diagnosis, were hospitalized for >30 days or started a continuous therapy of oral corticosteroids, because these events would preclude the evaluation of our main outcome. Data on medical and pharmacy services were obtained 1 year before cohort entry and during study follow-up.

Assessment of exposure

Follow-up time was divided into treatment episodes. A treatment episode was defined by the number of consecutive days a patient remained under the same treatment regimen. No minimum was required for the duration of a treatment period for inclusion in the analysis. The duration of a treatment period was based on the duration of the filled prescriptions and a delay was allowed of two times the duration of the prescription between renewals before considering that a patient had stopped his treatment. Each treatment episode was classified into one of seven treatment regimens for COPD based on filled prescriptions. The first regimen was formed of SABA and/or ipratropium bromide only. For all the other regimens, there was at least one medication added to the SABA and/or ipratropium bromide and one to three adjuvant therapies, i.e. theophyllines (aminophylline, theophylline or oxtriphylline in oral formulation); ICS (beclomethasone, triamcinolone, flunisolide, budesonide or fluticasone); LABA (salmeterol or formoterol); theophyllines and ICS concurrently; ICS and LABA concurrently; and theophyllines, ICS and LABA concurrently. Patients remained in a treatment episode until they added another medication to their treatment regimen, switched, stopped a medication or reached the end of the study follow-up. Patients could thus contribute to more than one episode of treatment during the study follow-up. An example of the assessment of exposure for a patient during the entire follow-up is presented.

Outcome

The main outcome was the rate of moderate to severe COPD exacerbations, i.e. the number of exacerbations divided by the person-days during episodes of a specific COPD treatment regimen. An episode was defined by the number of consecutive days a patient remained under the same treatment regimen. An exacerbation was defined either as a filled prescription of oral corticosteroids, a visit to an emergency department for COPD or a hospitalization for COPD (admission or principal ICD-9 codes equal to 491.x, 492.x or 496). Only a single exacerbation was considered if more than one marker of exacerbation [prescription of oral corticosteroids, Emergency Department (ED) visits for COPD or hospitalizations for COPD] occurred within a period of 15 days.

Potential confounders

Potential confounders included patient’s age at cohort entry, gender, socio-economic status [poor (receiving social assistance or receiving Guaranteed Income Supplement added to the Old Age Security Pension) vs. others] and the calendar year of cohort entry (to adjust for prescribing habits that may vary over time). Potential confounders measured in the year prior to cohort entry were also included, such as a medication-based comorbidity score and a medical visit-based continuity of care score. Markers of COPD severity included the number of prescriptions of oral corticosteroids, ED visits for COPD and hospitalizations for COPD in the year prior to cohort entry, as well as a medical visit with a respiratory physician, the average daily dose of SABA and ipratropium bromide (one dose equals two inhalations) and the number of prescriptions of antibiotics for COPD filled in the 3 months prior to each specific treatment episode.

Statistical analysis

The crude rate of moderate to severe COPD exacerbations was estimated for all treatment regimens. Poisson regression models were also performed to estimate the adjusted rate ratios of moderate to severe COPD exacerbations comparing patients who had regimens with same number of adjuvant therapies: (i) theophyllines vs. ICS; (ii) theophyllines vs. LABA; and (iii) theophyllines plus ICS vs. LABA plus ICS. Treatment episodes with theophyllines plus LABA were excluded from the comparison because of the small number of patients under this treatment regimen. All comparisons were performed twice, once among all COPD patients and secondly among patients who had three exacerbations of COPD or more in the year prior to cohort entry. This stratification was done to be coherent with the Canadian guidelines, which recommend regular use of ICS only in patients with moderate to severe COPD who have three or more acute exacerbations per year. Other Poisson regression models were used to obtain adjusted rate ratios for COPD exacerbations in association with the average daily doses of theophyllines and ICS. All potential confounder variables were included in the models and analyses were carried out using the SAS system version 8.2 (SAS Institute Inc., Cary, NC, USA).

Filed under: Pharmacokinetics