Abstract: Although controller therapies are currently recommended for patients who have persistent asthma, a number of studies indicate that these therapies do not adequately control asthma in a substantial number of these patients. This observation, combined with the potential risk of adverse effects with corticosteroids, supports the conclusion that controller therapies are not appropriate for all patients. However, some patients who do not respond to one type of controller therapy will respond to another, which suggests that we might consider targeting specific medications to select patients. There is increasing evidence that certain biomarkers may be useful in guiding therapy. For example, levels of sputum eosinophils have been shown to predict which patients are at increased risk for deterioration of asthma when inhaled corticosteroids are withdrawn.
The argument that the routine use of controller therapy is not indicated for all patients who have asthma is based on 3 points. First, in my opinion, controller therapy is not effective for many of the patients for whom such therapy is currently recommended.
Second, there is evidence that controller therapy is often unnecessary to achieve the goals of therapy in these patients. Finally, some of the asthma therapies we are currently using are potentially unsafe when used at high doses and for long periods.
Here I will present the evidence that supports these points, after a brief review of the goals of asthma therapy and the current recommendations for the use of controller therapy.
The goals of therapy
The goals of asthma therapy stipulated by the Global Initiative for Asthma report1 are consistent with those in the National Asthma Education and Prevention Program (NAEPP) guidelines (Table 1).2,3 The goals are to:
Minimize, or even obliterate, chronic symptoms, including those that occur at night.
Reduce the frequency of asthma exacerbations.
Keep patients out of the emergency department.
Reduce or eliminate the use of bronchodilators needed for breakthrough symptoms.
Ensure that patients can participate in normal activities, including exercise.
Ensure that patients have normal or nearly normal lung function and have modest variation in peak expiratory flow (PEF) (although the latter is probably more relevant for guidelines than it is for clinical practice).
Minimize or avoid adverse effects from medication.
Types of therapies
Asthma medications are generally categorized as controllers or relievers. Controllers are taken daily on a long-term basis to try to achieve and maintain a baseline level of control in persons who have persistent asthma. These include anti-inflammatory agents, such as corticosteroids (inhaled or, for more severe asthma, systemic corticosteroids); agents that modify the action or the synthesis of leukotrienes; anti-IgE agents; and long-acting bronchodilators.
The reliever medications used to treat acute symptoms include the rapid-onset bronchodilators--the short-acting ß2-agonists and short-acting anticholinergics--and systemic corticosteroids, which are administered to achieve rapid effects in patients with deteriorating asthma.
Who requires controller therapy?
The pharmacotherapy for asthma is generally pursued in a stepwise approach. The NAEPP guidelines do not recommend the use of controller medications for patients who have mild intermittent asthma.3 These patients can use a bronchodilator, as needed, if symptoms occur.
Controller therapy on a regular basis is reserved for patients with persistent symptoms, which are those occurring more than twice a week or nocturnal symptoms occurring more than twice monthly, or for patients who have PEF variability of 20% or more. For those patients, low-dose inhaled corticosteroids are recommended, although other medications, such as the leukotriene modifiers, are used as alternatives. The dosage can be increased or other medications added as the severity of the patient's asthma increases.
Controller therapy is not recommended for patients with mild intermittent asthma because it does not appear to produce clinically meaningful improvements in these patients, although studies have shown that controller agents suppress markers of inflammation, such as sputum eosinophils, exhaled nitric oxide, and methacholine responsiveness.4-7 This may reflect the fact that it is very difficult to demonstrate benefits of medications in patients with very mild disease.
Large surveys have been conducted to stratify patients into categories of asthma severity (Table 2). In the United States, about 19 million persons currently have asthma, and nearly 40% of them--7,600,000--have mild intermittent asthma.8 If we accept these data at face value, we might conclude that up to 40% of persons with asthma are not candidates for long-term controller therapy. However, it must be acknowledged that patients generally are not good at classifying their own asthma.
One reason not to use controller agents in those with mild asthma is that these therapies can be expensive. Inhaled corticosteroids cost between $500 and $600 a year for a patient. That would mean an incremental $4.2 billion in drug costs. I think that is a strong rationale for not prescribing medications for this group of patients.
Efficacy of controller therapy
In evaluating the indications for controller therapy, it is important to ask whether the patients for whom such therapy is currently recommended are really benefiting. Several studies suggest that many patients are not receiving the expected benefits.
Malmstrom and associates,9 for example, compared beclomethasone (200 µg twice daily), montelukast (10 mg once daily), and placebo in persons with mild to moderate asthma. At study entry, their forced expiratory volume in 1 second (FEV1) was 55% to 80% of predicted and they had bronchodilator responsiveness. Compared with placebo, both beclomethasone and montelukast resulted in significant improvements in FEV1. However, it is important to recognize that many of the patients--up to 25% to 30%--received no significant clinical benefit from beclomethasone or montelukast. When the analysis is based on a less than 10% improvement in FEV1, the percentage is even higher--greater than 50%.
It is reasonable to ask whether a better response might be achieved if patients were either given higher dosages of inhaled corticosteroids or given them for longer periods. Szefler and colleagues10 of the Asthma Clinical Research Network (ACRN) addressed this question. They investigated the variable nature of the response to inhaled corticosteroids in persons with mild to moderate asthma. The participants had not been using inhaled corticosteroids, and they had bronchodilator responsiveness, so they had the capacity to improve their lung function.
The participants were randomized to receive either beclomethasone or fluticasone, in escalating doses over 6-week periods, starting at low doses and reaching high doses at the end of the trial. Cortisol suppression was assessed with an overnight cortisol stimulation test.
Only about one third of the participants had a significant improvement in FEV1 (greater than 15% improvement).10 FEV1 did not improve in the later part of the trial, despite the fact that the inhaled corticosteroid dosages were escalated. However, there was a clear worsening of adrenal axis suppression with higher dosages. Thus, not only are some patients not getting any benefits from inhaled corticosteroids, but also they are experiencing the systemic effects of higher dosages of these agents.
Another important study by Szefler and colleagues11 evaluated children aged 6 to 17 years who had mild to moderate asthma. The children received fluticasone or montelukast for 8 weeks and then crossed over to the other treatment. Response to therapy was defined as an improvement in FEV1 of 7.5% or more.
The study found that 23% of the children responded to fluticasone only, 5% responded to montelukast only, and 17% responded to both.11 The majority, 55%, did not respond to either medication (Figure 1). These results raise the question of whether these medications should be uniformly prescribed for this patient population. The results also demonstrate that some patients who do not respond to one type of controller therapy will respond to another, which suggests that we might consider targeting specific medications to specific patients.
It is important to note that FEV1 may not be the best way to measure clinical response, and there are other ways to measure asthma control. It is probably more appropriate to think about asthma control in a composite fashion. This is how patients tend to think about it--they integrate the frequency and severity of their symptoms with other relevant factors, such as activity limitations and frequency of exacerbations. The key question, then, is: How well do these controller medications achieve a composite level of asthma control on a regular basis in the patients for whom they are prescribed?
This question was addressed by a 1-year, randomized, double-blind, parallel group study, known as GOAL (Gaining Optimal Asthma controL).12 The study enrolled over 3000 nonsmokers who had a history of asthma and had demonstrated a bronchodilator response. At baseline, their asthma had not been well controlled for at least 2 of the 4 weeks of the run-in period. The participants were stratified into the following 3 groups on the basis of their use of inhaled corticosteroids when they entered the trial: corticosteroid-naive, low-dose corticosteroids, and moderate-dose corticosteroids.
The participants were randomized to receive either the combination of fluticasone and salmeterol or an equivalent dosage of fluticasone. Those who entered the trial receiving low-dose inhaled corticosteroids or who were corticosteroid-naive started at the lower step of either the combination or fluticasone. Those who entered the trial at higher dosages of inhaled corticosteroids started at the middle step of therapy. These patients received escalating doses of inhaled corticosteroids until they had achieved total control of asthma or reached the highest dosage of either therapy.
Each assessment period was 12 weeks long. In the second phase of the study, patients were monitored for 1 year to assess whether asthma control could improve further with continued therapy.
The results were quite striking. In all 3 strata, the combination of salmeterol and fluticasone resulted in better asthma control than did fluticasone alone.12 However, even in the patients with mild asthma who had not been using inhaled corticosteroids--those who probably had the greatest likelihood of having a meaningful response--22% were unable to achieve the well-controlled status, even at higher dosages. The proportion of persons who did not achieve asthma control was even higher in the other groups.
In conclusion, our currently available controller agents are ineffective for many patients with asthma. A substantial portion of those with persistent asthma derive minimal, if any, benefit, at least as reflected by improvements in lung function. Moreover, the above- described GOAL study12 demonstrated that a significant proportion of patients do not achieve asthma control, as assessed by composite measures of control, when they receive inhaled corticosteroids with or without a long-acting ß2-agonist. These findings should be taken into consideration in deciding whether we should be prescribing these agents uniformly.
Is controller therapy necessary?
Another reason that we need to readjust the target population for controller agents is that, in my opinion, there are many patients for whom these agents are unnecessary. The ACRN recently addressed the question: Do patients who have mild persistent asthma require a daily controller therapy--that is, either an inhaled corticosteroid or a leukotriene receptor antagonist--to prevent the progressive loss of pulmonary function?13 This is often given as a rationale for providing such therapy.
The participants in this trial had persistent asthma, an FEV1 of at least 70% of predicted, and bronchodilator responsiveness or methacholine hyperresponsiveness, and they were taking ß2-agonists only. After 6 weeks of placebo, they were randomized to receive either continuous therapy with budesonide (200 µg twice daily), continuous therapy with zafirlukast (20 mg twice daily), or placebo for 12 months. In addition, every participant had access to open-label inhaled and oral corticosteroids, which they were instructed to use according to a symptom-based action plan.
Among the different treatment groups, there were no significant differences in the change in morning PEF.13 In fact, the outcomes were strikingly similar. There also was no difference in the number of episodes that warranted the use of prednisone.
Access to as-needed use of prednisone or inhaled corticosteroids does not appear to explain why the treatment groups were similar. The frequency of use of the symptom-based action plan was similar across the 3 groups. Importantly, most of the increased symptom episodes were ignored by patients.
Are there ways to identify the need for controller therapy on the basis of patient-specific phenotypic markers? Green and colleagues14 assessed the ability of sputum eosinophils, a measure of airway inflammation, to serve as an identifying factor in patients with moderate to severe asthma. The patients' inhaled corticosteroid dosages were titrated either according to British Thoracic Society guidelines,15 which are similar to the US guidelines,2,3 or according to the percentage of eosinophils in their sputum. The corticosteroid dosage was increased if the eosinophil percentage was greater than 3%. Otherwise, the dosage was either maintained or reduced if the eosinophil count was low for 2 consecutive measurements. If they had more symptoms and low sputum eosinophil counts, the ß2-agonists were increased.
Although the overall inhaled corticosteroid dosages were similar in the 2 treatment groups, the incidence of severe exacerbations was lower in those who had their corticosteroid dosages titrated according to sputum eosinophils.14 This indicates that this marker allowed specific allocation of higher dosages to those who needed it and thereby produced better asthma control compared with guideline-based therapy. The authors concluded that a strategy directed at maintaining a normal airway eosinophil count could reduce the incidence of asthma exacerbations.
To determine whether sputum eosinophil counts could be helpful in formulating a longer-term plan for patients receiving inhaled corticosteroids, we performed a retrospective analysis16 of the ACRN Salmeterol Or Corticosteroids Study (SOCS).17 SOCS examined patients with mild persistent asthma whose condition was stable at moderate dosages of inhaled corticosteroids and who had normal lung function and low PEF variability. They were randomized to either continue corticosteroids or switch to salmeterol or placebo, and they were observed for over 16 weeks.
One of the major findings of the SOCS trial was that compared with continued inhaled corticosteroids, a switch to placebo was associated with higher rates of asthma deterioration and higher levels of markers associated with airway inflammation, such as exhaled nitric oxide, sputum eosinophils, and airway hyperresponsiveness.17
We analyzed these data to determine whether these measurements--both at the time of randomization and shortly after inhaled corticosteroids were discontinued--could predict whose asthma would be poorly controlled or remain stable with placebo.16 Our analysis, which used a receiver operating characteristic curve approach to determine the ability of each of these markers to pre- dict subsequent deterioration, suggested that sputum eosinophils and, to a slightly less significant extent, sputum eosinophil cationic protein were strong predictors of asthma deterioration when corticosteroids were stopped.
In fact, the change in sputum eosinophil counts over the 2 weeks after inhaled corticosteroid discontinuation was highly sensitive and moderately specific for deterioration occurring over 14 weeks.16 About half the participants had shifts to low sputum eosinophil counts over the 2-week period. In these persons, the risk of asthma deterioration while receiving placebo, over 14 weeks of continued observation, was identical to that in the group that continued to receive inhaled corticosteroids. Neither exhaled nitric oxide nor methacholine hyperresponsiveness level could predict which patients would remain stable.
The implications of these findings are significant. Given that about 50% of the persons in the SOCS trial17 would qualify for discontinuing inhaled corticosteroids, and extrapolating to the 7 million Americans who have mild to moderate asthma, the savings in drug therapy approach billions of dollars. This does not even take into account the additional benefits of decreased toxicity from unnecessary corticosteroid treatment.
The potential adverse effects of systemic corticosteroid exposure have been well established and include osteoporosis,18 cataracts,19 and glaucoma.20
Israel and associates21 monitored 3 cohorts of women--those who were using no inhaled corticosteroids, those using moderate doses of inhaled corticosteroids, and those using high doses. They standardized the corticosteroid therapy the patients were receiving and observed them closely for 3 years to collect information such as dietary calcium intake and the numbers of oral corticosteroid bursts that were needed.
At the end of that 3-year study period, it was very clear that there was a puff-for-puff relationship between the amount of inhaled corticosteroids that these women were using and the loss of bone density measured at the trochanter (Figure 2).21 The change in bone density may seem numerically small, but it is important to extrapolate this loss of bone density over the life of a woman who, for example, might take moderate to high doses of corticosteroids at age 30 and have increased bone loss after menopause.
The incremental rate of bone loss imparted by continuous use of inhaled corticosteroids would confer a 2-fold increased risk of hip fracture in that woman. This is a risk that we have to consider when prescribing therapy.
With respect to other controller therapies, there may be an increased risk of mortality in a very small number of patients taking long-acting ß2-agonists,22 and increased liver enzymes have been reported in a small number of patients taking the 5-lipoxygenase inhibitor zileuton.23
In my opinion, controller therapy is not warranted for all patients with asthma. These types of medications are ineffective in many patients, according to lung function and composite measures of asthma control. These agents may be unnecessary in patients with mild persistent disease and in patients who have low levels of certain markers of inflammation, such as sputum eosinophils. These therapies are relatively expensive, and some of them are toxic at higher doses. I think the challenge before us, as asthma care practitioners, is to define the appropriate patient populations to ensure that we administer these medications to the patients who can benefit from them.
Global Initiative for Asthma (GINA). Global strategy for asthma management and prevention: NHLBI/WHO Workshop Report. Bethesda, Md: National Institutes of Health, National Heart, Lung, and Blood Institute; 2002. Publication 02-3659.
National Asthma Education and Prevention Program. Guidelines for the diagnosis and management of asthma: expert panel report 2. Bethesda, Md: National Institutes of Health, National Heart, Lung, and Blood Institute; 1997. Publication 97-4051.
National Asthma Education and Prevention Program report: guidelines for the diagnosis and management of asthma update on selected topics--2002 [published correction appears in
J Allergy Clin Immunol.
2003; 111: 466].
J Allergy Clin Immunol.
2002; 110 (suppl): S141-S219.
Kharitonov SA, Yates DH, Barnes PJ. Inhaled glucocorticoids decrease nitric oxide in exhaled air of asthmatic patients.
Am J Respir Crit Care Med.
1996; 153: 454-457.
Silkoff PE, McClean P, Spino M, et al. Dose-response relationship and reproducibility of the fall in exhaled nitric oxide after inhaled beclomethasone dipropionate therapy in asthma patients.
2001; 119: 1322-1328.
Lim S, Jatakanon A, John M, et al. Effect of inhaled budesonide on lung function and airway inflammation. Assessment by various inflammatory markers in mild asthma.
Am J Respir Crit Care Med.
Deykin A, Halpern O, Massaro AF, et al. Expired nitric oxide after bronchoprovocation and repeated spirometry in patients with asthma.
Am J Respir Crit Care Med.
1998; 157: 769-775.
The state of asthma in America: two landmark surveys. Available at: www.asthmainamerica. com. Accessed July 22, 2005.
Malmstrom K, Rodriguez-Gomez G, Guerra J, et al. Oral montelukast, inhaled beclomethasone, and placebo for chronic asthma. A randomized, controlled trial.
Ann Intern Med.
Szefler SJ, Martin RJ, King TS, et al. Significant variability in response to inhaled corticosteroids for persistent asthma.
J Allergy Clin Immunol.
Szefler SJ, Phillips BR, Martinez FD, et al. Characterization of within-subject responses to fluticasone and montelukast in childhood asthma.
J Allergy Clin Immunol.
2005; 115: 233-242.
Bateman ED, Boushey HA, Bousquet J, et al. Can guideline-defined asthma control be achieved? The Gaining Optimal Asthma ControL Study.
Am J Respir Crit Care Med.
2004; 170: 836-844.
Boushey HA, Sorkness CA, King TS, et al. Daily versus as-needed corticosteroids for mild persistent asthma.
N Engl J Med.
2005; 352: 1519-1528.
Green RH, Brightling CE, McKenna S, et al. Asthma exacerbations and sputum eosinophil counts: a randomized controlled trial.
British Thoracic Society. The British guidelines on asthma management.
Deykin A, Lazarus SC, Fahy JV, et al. Sputum eosinophil counts predict asthma control after discontinuation of inhaled corticosteroids.
J Allergy Clin Immunol.
2005; 115: 720-727.
Lazarus SC, Boushey HA, Fahy JV, et al. Long-acting ß
-agonist monotherapy vs continued therapy with inhaled corticosteroids in patients with persistent asthma: a randomized controlled trial.
Ledford D, Apter A, Brenner AM, et al. Osteoporosis in the corticosteroid-treated patient with asthma.
J Allergy Clin Immunol.
1998; 102: 353-362.
Garbe E, Suissa S, LeLorier J. Association of inhaled corticosteroid use with cataract extraction in elderly patients [published correction appears in
1998; 280: 539-543.
Garbe E, LeLorier J, Boivin JF, Suissa S. Inhaled and nasal glucocorticoids and the risks of ocular hypertension or open-angle glaucoma.
1997; 277: 722-727.
Israel E, Banerjee TR, Fitzmaurice GM, et al. Effects of inhaled glucocorticoids on bone density in premenopausal women.
N Engl J Med.
Castle W, Fuller R, Hall J, Palmer J. Serevent nationwide surveillance study: comparison of salmeterol with salbutamol in asthmatic patients who require regular bronchodilator treatment.
Lazarus SC, Lee T, Kemp JP, et al. Safety and clinical efficacy of zileuton in patients with chronic asthma.
Am J Manag Care.
1998; 4: 841-848.