SMART
PubMed • Full text • PDF
Clinical Question
In asthmatics, does the addition of the long-acting ß-agonist salmeterol to asthma therapy increase respiratory-related adverse events and mortality?
Bottom Line
Salmeterol increases the risk of respiratory-related deaths, particularly among African American patients and those not using an inhaled corticosteroid.
Major Points
Long-acting beta-agonists (LABAs) such as salmeterol have since the 1990s been mainstays of therapy for persistent asthma. This trial was designed after observational studies suggested an increase in asthma-related mortality despite optimal medical therapy, raising the question that perhaps some deaths were related to complications of therapy.
The Salmeterol Multicenter Asthma Research Trial (SMART) randomized over 26,000 adolescent and adult patients with asthma to salmeterol or placebo in addition to standard asthma therapy. There was no difference in the composite primary outcome of respiratory-related deaths or life-threatening events with salmeterol. However, salmeterol was associated with significant increases in secondary outcomes such as asthma-related deaths or life-threatening events, deaths from asthma, and respiratory-related deaths. Subgroup analyses revealed that this increased morbidity and mortality was most pronounced among African Americans and those not taking an inhaled corticosteroid (ICS).
In large part due to the results of SMART, the 2007 NHLBI ECS 3 guidelines recommend adding a LABA to ICS therapy as step 3 of treatment for persistent asthma (level of evidence A). In addition, the FDA provides a black box warning on LABA preparations because of their association with asthma-related deaths.
Design
- Multicenter, double-blind, parallel-group, randomized, placebo-controlled trial
- N=26,355 asthmatics
- Salmeterol (n=13,176)
- Placebo (n=13,179)
- Setting: 6,163 sites
- Enrollment: 1996-2003
- Follow-up: 28 weeks (median)
- Analysis: Intention-to-treat
Population
Inclusion Criteria
- Age ≥12 years
- Diagnosis of asthma
- Receiving ≥1 prescription asthma medication
Exclusion Criteria
- Prior use of LABA
- Use of ß-blockers
- Pregnancy or lactation
- Significant systemic disease
Baseline Characteristics
- Mean age: 39.2 years
- Female: 64%
- Race/ethnicity:
- Caucasian: 71%
- African-American: 18%
- Hispanic: 8%
- Asian: 1%
- Other: 2%
- PEF: 355 L/min (82% predicted)
- Duration of asthma: 16.3 years
- Emergency visits in prior year: 26%
- Hospitalizations in prior year: 8%
- Nocturnal symptoms: 61%
- Inhaled corticosteroid use: 47%
Interventions
- Randomized to salmeterol xinafoate MDI 42µg BID vs. placebo MDI BID for 28 wks
- Initial office visit for screening and randomization; subsequent follow-up via telephone every 4 wks
- All patients were also given a supply of albuterol
Outcomes
Comparisons are salmeterol vs. placebo. P values weren't provided by the authors.
Primary Outcome
- Respiratory-related deaths or life-threatening events
- 50 (<1%) vs. 36 (<1%), RR 1.39 (95% CI 0.91 to 2.14)
- African-Americans: 20 (<1%) vs. 5 (<1%), RR 4.10 (95% CI 1.54 to 10.90)
- Caucasians: 29 (<1%) vs. 28 (<1%), RR 1.05 (95% CI 0.62 to 1.76)
Secondary Outcomes
- All-cause mortality
- 42 (<1%) vs. 36 (<1%), RR 1.30 (95% CI 0.82 to 2.06)
- African-Americans: 12 (<1%) vs. 7 (<1%), RR 1.69 (95% CI 0.67 to 4.28)
- Caucasians: 29 (<1%) vs. 22 (<1%), RR 1.32 (95% CI 0.76 to 2.30)
- Asthma-related deaths or life-threatening events
- 37 (<1%) vs. 22 (<1%), RR 1.71 (95% CI 1.01 to 2.89)
- African-Americans: 19 (<1%) vs. 4 (<1%), RR 4.92 (95% CI 1.68 to 14.45)
- Caucasians: 17 (<1%) vs. 16 (<1%), RR 1.08 (95% CI 0.55 to 2.14)
- Deaths from asthma
- 13 (<1%) vs. 3 (<1%), RR 4.37 (95% CI 1.25 to 15.34)
- African-Americans: 7 (<1%) vs. 1 (<1%), RR 7.26 (95% CI 0.89 to 58.94)
- Caucasians: 6 (<1%) vs. 1 (<1%), RR 5.82 (95% CI 0.70 to 48.37)
- Respiratory-related deaths
- 24 (<1%) vs. 11 (<1%), RR 2.16 (95% CI 1.06 to 4.41)
- African-Americans: 8 (<1%) vs. 2 (<1%), RR 3.89 (95% CI 0.82 to 18.26)
- Caucasians: 16 (<1%) vs. 7 (<1%), RR 2.29 (95% CI 0.94 to 5.56)
- Combination of death or life threatening experiences
- 70 (<1%) vs. 59 (<1%), RR 1.19 (95% CI 0.84 to 1.68)
- African-Americans: 24 (1%) vs. 11 (<1%), RR 2.17 (95% CI 1.06 to 4.41)
- Caucasians: 44 (<1%) 44 vs. (<1%), RR 1.01 (95% CI 0.67 to 1.53)
- Hospitalizations
- 469 (4%) vs. 420 (3%), RR 1.11 (95% CI 0.97 to 1.26)
- African-Americans: 102 (4%) vs. 77 (3%), RR 1.29 (95% CI 0.96 to 1.72)
- Caucasians: 323 (3%) vs. 317 (3%), RR 1.02 (95% CI 0.88 to 1.19)
Subgroup Analysis
For the primary outcome.
- ICS at baseline
- Yes: 23 (<1%) 19 (<1%), RR 1.21 (95% CI 0.66 to 2.23)
- No: 27 (<1%) 17 (<1%), RR 1.60 (95% CI 0.87 to 2.93)
Adverse Events
- Frequency of serious adverse events
- 4% in each group
- Time to first serious event causing discontinuation, in survival rate
- 95.6% vs. 96.2% (P=0.02)
Criticisms
Unable to tell if differences in African American group is from imbalances between baseline factors (e.g. use of controller medications, socioeconomics)
Funding
Multiple authors with connections to GlaxoSmithKline, the makers of Serevent.