Reviews/Evaluations

Fluticasone/Salmeterol (Advair™) Drug Utilization Review

I. Background

OMAP drug utilization trend data shows more than 30% increase in expenditures for asthma related medications during the year 2001. A newly formulated combination product, fluticasone/salmeterol (Advair™) marketed in March 2001, was identified as one of the major pharmacologic agents responsible for this rapid increase in asthma spending (Figure 1). Advair™ expenditures increased to $0.68 per member per month from March 2001 to March 2002.

Figure 1

Asthma is a relatively common reactive airway disease (RAD) that is associated with substantial morbidity, mortality, and resource utilization.(1) In 1999, it was estimated that 17 million Americans suffered from asthma. Healthcare costs of asthma management continue to increase. The direct costs include drug therapy, emergency room visits and inpatient admissions. Collectively, direct medical costs are approximately $3.6 billion annually. Indirect costs, associated with lost time at work are approximately $2.6 billion in 1990.(2)

As with many other chronic diseases, pharmacotherapy is important to reduce the risk of asthma related mortality, decrease disability and improve symptoms and quality of life. Inhaled corticosteroids (ICS) are considered the most effective therapy for long-term control of asthma. Their potent anti-inflammatory actions directly contribute to reductions in symptom severity, improvements in measures of airflow mechanics, diminution in airway hyperresponsiveness and preventing irreversible airway remodeling.(3) Both large-scale epidemiologic studies and randomized controlled clinical trials have consistently demonstrated that ICS significantly reduces asthma related morbidity and mortality.(4) Long-acting beta-agonists (salmeterol, formoterol) and leukotriene modifiers are useful adjuncts to anti-inflammatory therapy for long-term control of asthma symptoms.

II. Clinical Practice Guidelines

The National Asthma Education and Prevention Program (NAEPP) provides clinical practice guidelines for the diagnosis and management of asthma with emphasis on the need for appropriate medication use to manage both acute symptoms and to establish long-term control. The NAEPP Expert Panel identifies key areas of asthma management and uses a systematic review of research evidence to provide periodic updates for health care providers. The most recent update on selected topics was released in June 2002. The update addresses the role of combination therapy in patients with moderate persistent asthma.(1)

For adults and children over 5 years of age with moderate persistent asthma who are receiving ICS, the recommendations were revised to include the addition of long-acting inhaled beta2-agonists (LABA) to low-to-medium doses ICS for long-term control. Adding a leukotriene modifier or theophylline to low-to-medium doses ICS is an alternative option, but the evidence is not as compelling. For children younger than 5 years of age, adjunct therapy combinations have not been studied. The Expert Panel recommends two preferred options for treating moderate asthma for this age group: either the addition of LABA to a low-dose ICS or medium-dose ICS as monotherapy.(1)

These recommendations are based on evidence from several large controlled clinical trials that have consistently shown the addition of a LABA to low-to-medium doses of ICS achieves greater improvement in pulmonary function and asthma symptoms and reduction in supplemental use of beta2-agonists than ICS dose escalation alone in patients with mild or moderate persistent asthma.

III. Summary of Clinical Trials

Many large, multicenter, double-blinded, randomized controlled trials evaluated the clinical efficacy of LABA plus low-to-medium doses of ICS in patients with mild to moderate persistent asthma with or without history of prior ICS use.(5-17) Treatment period ranged from 12 weeks to 1 year. Patients were required to be at least 12 years of age with forced expiratory rate in one second (FEV1) around 40-85% at the time of randomization. Different combinations and at various doses of LABA (formoterol or salmeterol) and ICS (fluticasone, budesonide, beclomethasone or triamcinolone) were evaluated in these trials. Some of these studies included Advair™ as one of the comparative arms.(13) Outcome measures included changes from baseline to endpoint in pulmonary function [e.g. FEV1, morning/evening peak expiratory flow rate (PEF)], rate of exacerbation (i.e., requirement of inhaled short-acting beta2-agonists or oral or parenteral corticosteroid treatment), overall asthma symptom scores, and symptom free days or nights.

In summary, adding LABA to low-to-medium doses of ICS showed significantly better improvement in pulmonary function as evidenced by positive change in FEV1 and morning/evening PEF than doubling doses of ICS alone in patients with mild to moderate persistent asthma.(5-7,9-15) Rates of exacerbation had mixed results in these trials. While supplemental inhaled short-acting beta2-agonists use was reduced more significantly in LABA/ICS treated patients than doubling doses of ICS alone treated patients, the rate of oral or parenteral corticosteroid treatment requirement or exacerbation event reduction was not significantly different between combination therapy and higher doses of ICS.(5,7-13,15) There was a slight trend of lower rate of exacerbation in the combination treated patients.(5-8,10,11) Looking at other outcome measures, reduction in overall asthma symptom scores, increase of symptom free days or nights, and time to first exacerbation had all favored the combination therapy.(6,7,9-12,14)

For patients with mild asthmatic disease treated with short-acting beta-agonists alone, two clinical trials showed that initiation of maintenance therapy with salmeterol and ICS provided significantly greater improvement in mean change from baseline to endpoint in pre-dose FEV1 and PEF than initiation of maintenance therapy with ICS alone.(5,6) However, no statistically significant differences were noted among treatment groups with respect to nighttime awakenings due to asthma, supplemental albuterol use, time to the first severe exacerbation or rate of severe asthma exacerbation per patient per year in both studies.(5,6)

Two unpublished studies evaluated the clinical efficacy of Advair™in COPD patients were submitted to the FDA.(18) Treatment period in both trials was 24 weeks. Although patients with asthma were excluded, 54-55% of patients with airway reversibility were enrolled in these studies. In addition, the studies were limited to only patients with confirmed chronic bronchitis. The two study designs were similar. Patients were randomized into one of the four arms treatment: Advair™, monotherapy of inhaled salmeterol or inhaled fluticasone, or placebo.

Looking at the primary outcomes from both studies, Advair™ treated patients showed significantly better improvement in pre-dose and 2 hour post-dose FEV1 than placebo treated patients.(18) However, when results were compared with monotherapy of either salmeterol or fluticasone, Advair™ was not consistently better than either treatment. Subgroup analysis of primary efficacy endpoints indicated that patients with reversible disease obtained a greater treatment effect (63%) than patients with non-reversible disease in both studies. For the secondary endpoints of global assessment of symptoms and quality of life measurements by subjects, frequency and severity of COPD exacerbations, and percentage of withdrawals due to COPD exacerbations, no statistically significant difference was found between Advair™ and placebo, or monotherapy of either salmeterol or fluticasone. In the assessment of dyspnea, Advair 500/50 had a meaningful improvement in dyspnea compared with placebo and salmeterol, but Advair 250/50 did not.(18)

The FDA review of these two studies raised several issues regarding the efficacy and safety of Advair™ as long-term maintenance therapy in patients with COPD.(18) First, taken all the efficacy measures, the failure to demonstrate efficacy with secondary endpoints raised the concern regarding the clinical relevance of the FEV1 findings since the efficacy of Advair™ on airflow limitation did not translate into a clear clinical benefit. Second, the efficacy findings in these two trials may not be applicable to the COPD population as a whole because patients with non-reversible disease had only small treatment effect and the studied population consisted only patients with chronic bronchitis, which is only a sub-population of COPD, therefore not fully representative of COPD. Third, higher adverse events were noted in the Advair™ group compared to the placebo group. These two studies were not designed, nor were they of significant duration, to evaluate bone mineral density or ocular effects. Therefore, safety of Advair™ has only been established for short-term treatment.

Nevertheless, the FDA Pulmonary & Allergy Drugs Advisory Committee voted in favor to approve Advair™for the indication of long-term maintenance treatment of COPD (including emphysema and chronic bronchitis) with several recommendations, including a very stringent labeling restrictions regarding the difference between COPD and chronic obstructive bronchitis, a very specific language on the duration of therapy and the requirement of a dose escalation study in post-marketing trial.(19) At this time, the FDA has not approved Advair™for the treatment of COPD.

IV. Recommendations

Based on the review of NAEPP guidelines and clinical trials, it is recommended that patients with mild to moderate persistent asthma who remain symptomatic on ICS should add LABA rather than increase the ICS dose. Mild persistent asthma is defined as daytime symptoms >2/week (<1x/day) or nighttime symptoms >2x/month and predicted FEV1>80%. Salmeterol/fluticasone (Advair™) can be considered as a drug option as it includes both LABA and ICS. For patients with mild persistent asthma without prior treatment of ICS, there is no clear evidence to support additional benefit gained if they were initiated on combination of ICS/LABA rather than ICS alone for long-term control. Therefore, for these patients, initiation of ICS monotherapy is recommended. Addition of LABA should be considered later if patient remains symptomatic.

Lastly, Advair™ is not approved for COPD patients at this time. The unpublished studies reviewed by the FDA showed marginal differences in clinical efficacy of this product compared to monotherapy with either salmeterol or fluticasone. In addition, the safety of long-term ICS has not been extensively studied. Therefore, the use of Advair™ in COPD patients should be discouraged.

V. Prescription Claims Evaluation

The cost of Advair and its individual components (salmeterol, fluticasone) are shown in Table 1 below. Depending on the dose, the cost of Advair is approximately the same or less than that of individual salmeterol and fluticasone inhalers.

Therefore, if used appropriately, Advair represents a convenient and cost-effective agent for the treatment of moderate to severe persistent asthma. However, there are some concerns that combination therapy is being initiated prematurely in patients who might otherwise be adequately managed on ICS alone.

1. Goal: Characterize prior drug use patterns for asthma management among Oregon Health Plan (OMAP) fee for service (FFS) patients that have had prescriptions filled for Advair.
   
   
2. Methodology: Retrospective prescription drug claims analysis: 1/1/01 - 4/1/02
   
   
3. Population: Inclusion/Exclusion
   A. > 1 Rx for Advair™ between 1/01/01 and 04/01/02
   B. > 3 months of continuous eligibility prior to sentinel fill
   C. No managed care plan enrollment during 3 months prior to sentinel fill
   
   
4. Objectives
   A. Determine the proportion of new start Advair patients that had a fill for the following medications within the 90 days prior to sentinel fill.
   

   1. No medications
   2. Short-acting beta-agonists (SABA)

   a. Albuterol (Ventolin™, Proventil™, misc. generic products)
   b. Bitolerol (Tornalate™)
   c. Metaproterenol (Alupent™)
   d. Pirbuterol (Maxair™)
   e. Terbutaline (Brethine™)

   3. Inhaled corticosteriods (ICS)

   a. Beclomethasone dipropionate (Proventil™, Beclovent™)
   b. Budesonide (Pulmicort™)
   c. Flunisolide (Aerobid™)
   d. Fluticasone (Flovent™)
   e. Triamcinolone (Azmacort™)

   4. Long-acting beta-agonists (LABA)

   a. Salmeterol (Serevent™)
   b. Formoterol (Foradil™)

   5. Leukotriene modifiers (LM)

   a. Montelukast (Singulair™)
   b. Zarfirlukast (Accolate™)
   c. Zileuton (Zyflo™)

   6. Ipratropium bromide (Atrovent™, Combivent™) - (IPBR)

   7. Mast Cell Stabilizers (MCS)

   a. Cromolyn
   b. Nedocromil

   B. Determine what proportion of patients can be identified as having asthma versus other obstructive respiratory diseases (COPD, chronic bronchitis) as identified by ICD-9-CM diagnosis (ICD9 1-5) coding during the years 2000 and 2001

   1. Asthma: 493.XXX (extrinsic, intrinsic asthma; chronic obstructive asthma, asthma unspecified)

   2. COPD: 490.XXX (bronchitis unspecified), 491.XXX (chronic bronchitis), 492.XXX (emphysema), 494.XXX (bronchiectasis), 495.XXX (extrinsic allergic alveolitis), 496.XXX (chronic airway obstruction, not elsewhere clarified).

Results

A total of 2144 unique patients had filled a prescription for Advair during the study months. Demographics data is conveyed in Table 2. Of these patients, 54% (1148) maintained uninterrupted eligibility, without managed care enrollment, for 3 months prior to their first fill (sentinel fill) of Advair™. The mean age of this group was 53 years (SD 23). However, ages ranged from 3 to 99 years. Close to 70% of patients in this analysis were female.

The distribution of age is shown in Figure 2. A bimodal distribution can be appreciated. The first mode is around 10 to 15 years of age. A second, more broad, mode is apparent around 60.

Figure 2.

Examination of medical claims reveals 426 patients (37%) of members have a diagnosis code of Asthma (Table 3). The mean age for this group is 47 (SD=21). A total of 344 (33%) patients have diagnosis code associated with COPD. The mean age for this group is 61. Clinically, these conditions are not always mutually exclusive. This is reflected in the medical clams where 179 (16%) patients have been coded both with asthma and COPD.

Table 4 below shows the results of the prescription claims analysis. The combination totals convey the number of patients on the drug classes described previously described. They are not mutually exclusive and therefore do not add up to 100%. For example, a patient as being on an ICS and LABA also was permitted to be on a 3rd agent and still be counted for the dual combination.

A total of 351 (31%) patients had at least one ICS filled in the 90 days prior to their sentinel Advair fill. Only 233 (20%) of patients had filled a LABA prior to their sentinel fill. Of these patients, 167 (14% of the total) had a fill for both an ICS and a LABA prior to the sentinel fill. There were 200 patients (17%) who had filled a LM prior to their sentinel fill. More than 30% of patients had a fill for an IPBR containing product.

Almost half of all patients analyzed (43%) did not have a fill for an asthma/COPD controller medication. Table 5 shows a breakdown of short-acting beta-agonist usage in these patients. Of this group, 71.5% also had no SABA filled during the 90 days prior. The remaining 141 (28.5%) patients had a least 1 fill for a SABA during the 90-day window.

VI. Limitations

There are several methodological concerns that must first be considered before conclusions can be drawn. First, administrative databases are not designed for clinical evaluations of drugs and disease. Problems with miscoding, misbilling, and discontinuous enrollment contribute to incomplete capture of heath resource utilization data. Second, the dynamic nature of a Medicaid population necessitate that efforts to control for the entrance and loss of patients from the population be made. In attempts to ensure a stable cohort of patients, only members with at least 3 months of eligibility prior to their sentinel fill were evaluated. The inclusion criteria employed may have inadvertently selected out a population of patients that is systematically unique and therefore a threat to external validity. It is commonly acknowledged that the Medicaid recipients who qualify for Medicaid under the blind and disabled eligibility criteria are among the most consistent in their enrollment. Third, inhaled asthma products, like other drugs, are heavily sampled by the pharmaceutical industry. It is possible that these patients could be receiving appropriate therapy through sampled products rather than prescriptions filled at a pharmacy. Finally, it was assumed that having filled a prescription within the last 90 days prior to the AdvairTM would account for stable use by the patient.

VII. Conclusions

The results of this DUR suggest that a large proportion of patients who were started on combination Advair™ are likely suffering from COPD as evidence of the numerous medical claims for a COPD related condition, the advanced age of the cohort, and the substantial number of patients also receiving ipratropium therapy. Little data exists supporting the use Advair™for the management of COPD. Two unpublished trials submitted to the FDA do suggest marginal benefit with respect to increases in FEV1.(18) However, these pulmonary function test improvements failed to translate into statistical or clinical improvements in symptom scores or exacerbation rates. The most marked improvements were seen in those patients who were characterized at baseline has having reversible disease. It has been suggested that the benefit of this drug product in COPD patients, if any, is limited to patients with a significant reversible airflow component.(19)

It is also apparent that only 30% of patients had a prescription filled for an ICS in the 3 months prior to their first Advair™ fill. Clinical trial data have suggested that patients with mild asthmatic disease requiring no ICS gain no additional benefit from combination therapy compared to an ICS alone as initial treatment.(5,6) An even smaller proportion of patients (17%) were taking both and ICS and a LABA prior to the sentinel Advair™ fill. It is in these patients that a combined drug product could possibly enhance compliance and reduce overall drug acquisition costs.

The most striking result was that close to half of all patients were on no controller prior to their first fill of Advar™ There is little clinical data to suggest the superiority of dual therapy in patients not previously managed with a controller. It could be hypothesized that these patients were initiated on combination therapy for aggressive control of previously uncontrolled mild to moderate or severe persistent asthma. If so, the use of inhaled short-acting beta2-agonists (SABA) would potentially be inflated. However, only 29% of patients on no controller medication had a fill for a SABA (Table 5). Of this group 50% had regular monthly fills during the 90 days prior to their sentinel fill. Thus it seems unlikely, that this is the case. This group also could represent patients who are relatively new to the Medicaid program and were previously on combination therapy or being stepped up from medication that was not captured in the claims.

References

1. National Asthma Education and Prevention Program. Guidelines for the diagnosis and management of asthma - update on selected topics 2002: NIH; 2002.
2. Wang SW, Liu X, Wiener DJ, Sennett C, Bowers BW, Legorreta AP. Comparison of prevalence, cost, and outcomes of a combination of salmeterol and fluticasone therapy to common asthma treatments. American Journal of Managed Care 2001;7:913-22.
3. Spahn J, Covar RSDA. Asthma: addressing consistency in results from basic science, clinical trials, and observation experience. Journal of Allergy & Clinical Immunology 2002;106:S490-S502.
4. Stempel DA, Kelly HW. Approaching value in asthma management: the need to integrate clinical and economic research with basic science. Journal of Allergy & Clinical Immunology 2002;109:S503-S510.
5. O'Byrne P, Barnes PJ, Rodriguez-Roisin R et al. Low dose inhaled budesonide and formoterol in mild persistent asthma: the OPTIMA randomized trial. American Journal of Respiratory & Critical Care Medicine 2001;164:1392-1397.
6. Pearlman DS, Stricker W, Weinstein S et al. Inhaled salmeterol and fluticasone: a study comparing monotherapy and combination therapy in asthma. Ann Allergy Asthma Immunol 1999;82(3):257-265.
7. Pauwels RA, Lofdahl CG, Postma DS et al. Effect of inhaled formoterol and budesonide on exacerbations of asthma. New England Journal of Medicine 1997;337:1405-1411.
8. Matz J, Emmett A, Rickard K, Kalberg C. Addition of salmeterol to low-dose fluticasone versus higher-dose fluticasone: an analysis of asthma exacerbations. Journal of Allergy & Clinical Immunology 2001;107:783-9.
9. Condemi JJ, Goldstein S, Kalberg C, Yancey S, Emmett A, Rickard K. The addition of salmeterol to fluticasone propionate versus increasing the dose of fluticasone propionate in patients with persistent asthma. Salmeterol Study Group. Annals of Allergy, Asthma, & Immunology 1999;82:383-9.
10. Murray JJ, Church N, Anderson WH et al. Concurrent use of salmeterol with inhaled corticosteroids is more effective than inhaled corticosteroid dose increases. Allergy and Asthma Proceedings 1999;20:173-180.
11. Kelson SG, Church NL, Gilman SA et al. Salmeterol added to inhaled corticosteroid therapy is superior to doubling the dose of inhaled corticosteroids: a randomized clinical trial. Journal of Asthma 1999;36:703-715.
12. Woolcock AJ, Lundback B, Ringdal N, Jacques LA. Comparison of addition of salmeterol to inhaled steroids with doubling of dose of inhaled steroids. American Journal of Respiratory & Critical Care Medicine 1996;153:1481-81.
13. Jenkins C, Woolcock AJ, Saarelainen P, Lundback B, James MH. Salmeterol/fluticasone propionate combination therapy 50/250 mcg twice daily is more effective than budesonide 800 mcg twice daily in treating moderate to severe asthma. Respiratory Medicine 2000;94:715-23.
14. Baraniuk J, Murray JJ, Nathan RA et al. Fluticasone alone or in combination with salmeterol vs triamcinolone in asthma. Chest 1999;116:625-32.
15. van Noord JA, Schreurs AJ, Mol SJ, Mulder PG. Addition of salmeterol versus doubling the dose of fluticasone propionate in patients with mild to moderate asthma. Thorax 1999;54:207-12.
16. Greening AP, Ind PW, Northfield M, Shaw G. Added salmeterol versus higher-dose corticosteroid in asthma patients with symptoms on existing inhaled corticosteroid. Lancet 1994:219-224.
17. Nelson HS, Busse WW, Kerwin E et al. Fluticasone propionate/salmeterol combination provides more effective asthma control than low-dose inhaled corticosteroid plus montelukast. Journal of Allergy & Clinical Immunology 2000;106:1088-95.
18. FDA Medical Officer Review: NDA#21-077/SE1-003 Advair Diskus. http://www.fda.gov/ohrms/dockets/ac/02/briefing/3830b1_02_Glaxo-ADVAIR.pdf.
19. FDA Pulmonary & Allergy Drugs Advisory Committee Minutes. http://www.fda.gov/ohrms/dockets/ac/02/minutes/3830m1.pdf.