Prof. G. W. Canonica Allergy and Respiratory Diseases Clinic, DIMI University of Genoa Pad. Maragliano L.go R. Benzi 10 16132 Genoa Italy
The objective of the study is to assess the efficacy of the nonsedating antihistamine, desloratadine, in the treatment of allergic rhinitis (AR). A search of MEDLINE, EMBASE, LILACS, and CINAHL databases was undertaken from January, 1966 to May, 2006. Double-blind, randomized, controlled studies of desloratadine in the treatment of AR in adult patients were carried out. The measured outcomes included the total symptoms score, the total nasal symptoms score, nasal airflow, and inflammatory markers (nasal eosinophils, nasal interleukin-4). The analysis included the calculation of standardized mean difference (SMD). A total of 57 studies were analyzed, and 13 randomized, double-blind, controlled trials were included in the meta-analysis. The trials included 3108 subjects who had completed studies involving desloratadine. There was significant heterogeneity among the study results, because of differing study methodologies. Desloratadine was associated with significant reductions in total symptoms scores (SMD −1.63; 95% CI −2.75 to −0.51; P = 0.004) and total nasal symptoms score (SMD −0.66; 95% CI −0.91 to −0.42; P < 0.001), when compared with placebo. Analysis of objective data on nasal blockage demonstrated a significant improvement in nasal airflow with desloratadine, when compared with placebo (SMD 0.32; 95% CI 0.10 to 0.55; P = 0.005). A benefit favoring desloratadine over placebo in terms of nasal eosinophil levels was also noted in the analysis. This meta-analysis confirms the reduction of AR symptoms and improvement in nasal airflow seen in individual studies of desloratadine. Objective improvements in nasal airflow, total symptoms, and total nasal symptoms seen with desloratadine are supported by Ia evidence.
Allergic rhinitis (AR) is a common immune-mediated disorder that has a prevalence of between 9% and 16% in the United States and up to 28.9% in a survey of countries worldwide (1, 2). In Italy, for instance, recent data suggest a prevalence up to 35.2% in pediatric subjects, which has increased by 5% in the last 5 years (3). As such, AR has a large healthcare burden, with annual total costs of about €3 billion in Europe and at least $5.6 billion in the United States (4). These figures include both direct costs (healthcare resource utilization and drug costs) and indirect costs (loss of productivity), with the latter accounting for the majority of the total cost.
Allergic rhinitis occurs because of the cross-linking of IgE by allergens within the upper respiratory tract. Allergen activation of IgE is followed by a complex local and systemic cascade of events, which includes immediate release of inflammatory mediators, including histamine, from mast cells in the nasal mucosa, neurovascular inflammation, and upregulation of Th2-type cytokines [e.g., interleukin-4 (IL-4)] in the circulation (5). These events are accompanied by the increased trafficking of cells, such as, eosinophils to the site of allergic activation, aided by the expression of adhesion molecules. The symptoms caused by this allergic inflammation include sneezing, rhinorrhea, nasal pruritus, and nasal congestion; allergic conjunctivitis often accompanies AR. Allergic rhinitis is classified either according to the chronicity of symptoms as per the ARIA guidelines (intermittent: ≤4 days per week or ≤4 weeks per year, persistent: >4 days per week and >4 weeks per year) or traditionally according to the seasonality of symptoms (seasonal or perennial AR) (6). The ARIA classification is favored by many, as it has been validated in everyday practice and concentrates more on the patient's symptoms than the time of year in which they occur (7).
Given the high prevalence and consequent significant healthcare burden attributable to AR, multiple guidelines for management are available (1). In a series of collaborations, EAACI and ARIA have devised guidelines for the treatment of AR, which focus on a stepwise approach to management (8, 9). Nonsedating antihistamines are a principal first-line therapy for AR; older first generation agents are not recommended by any of the major professional therapeutic guidelines, because of their adverse event profiles, particularly in terms of sedation (1). Many nonsedating antihistamines are clinically available and in individual studies have shown their effectiveness in the treatment of AR symptoms, particularly sneezing, watery rhinorrhea, and nasal itching. The efficacy of second-generation antihistamines against nasal congestion has been a matter of debate with heterogeneous results seen across the class.
Desloratadine is a nonsedating antihistamine that first became available in 2001 for the treatment of AR; desloratadine is currently approved for the treatment of AR irrespective of duration (intermittent/persistent) or seasonality (seasonal/perennial) of symptoms (10). Since its launch, a sizeable number of placebo-controlled trials demonstrating the efficacy of desloratadine in AR have been reported. Results also suggest that desloratadine is associated with symptomatic and particularly objective improvements (e.g., nasal airflow) in nasal congestion (11). It has been argued that such effects on nasal congestion may be due to systemic effects of desloratadine on mediators of allergic inflammation (12). Although the availability of multiple placebo-controlled trials regarding desloratadine is useful, higher degrees of validation exist. It is widely accepted that the highest standard of evidence-based medicine (type Ia) comes from well-designed formal meta-analysis of data from multiple studies (13). Such useful studies are becoming more frequent in the field of clinical allergy (14, 15). In an effort to assess the robustness of the activity of desloratadine in AR, we performed a meta-analysis of desloratadine data in terms of clinical symptom control (including nasal congestion) and objective measures on nasal airflow. In addition, we analyzed the effects of desloratadine on allergic inflammatory mediators in clinical trials of subjects with AR treated with desloratadine at clinically relevant doses.
Four electronic databases (MEDLINE, EMBASE, LILACS, and CINAHL) were searched from 1966 to May 31, 2006, for randomized placebo-controlled and double-blind trials investigating efficacy of desloratadine in patients with AR, with MeSH headings, and text words. We searched also for any additional study mentioned in the references of identified publications, including previous relevant meta-analyses, and narrative reviews. Abstracts of relevant meetings were also searched.
Two authors conducted independent search strategies. The first MEDLINE search strategy retrieved citations containing the subject heading desloratadine (limited to the publication types: clinical trial and meta-analysis) or the text words antihistamines, active metabolite or Aerius. The second MEDLINE search strategy retrieved citations containing the subject heading desloratadine combined with exploded subject headings describing allergic disease (rhinitis, hay fever, and rhinoconjunctivitis) or text words describing the efficacy of desloratadine in AR appearing in close proximity (desloratadine, allergy, rhinitis, and efficacy) or those focused to the target population (humans). We limited citations from the second search to randomized, controlled trials using a maximally sensitive strategy (16). We modified these searches for other databases. We screened reference lists from all retrieved articles and from recent review articles to identify additional studies. There were no language restrictions.
Study selection and characteristics
We included parallel-group, randomized, controlled, and double-blinded trials. Patients had to have a history of AR with or without allergic asthma and/or conjunctivitis, in which the causal allergen was identified; IgE sensitization was proved by prick test and/or specific IgE assays. Desloratadine had to be delivered by oral route and all doses and all durations of treatment were considered. Those trials of desloratadine, which included patients with allergic asthma or conjunctivitis were considered only if the results for subjects with rhinitis were separately analyzed. Trial eligibility was determined from the full text format of the retrieved study by two authors and checked by the principal investigator. The observed percentage agreement between the investigators for the assessment of inclusion was calculated using the κ test (17, 18).
Assessment of validity
Methodological information relevant to the assessment of internal validity was method of allocation, generation and concealment of randomization, blinding of caregivers/outcome assessors, number of and reasons for postrandomization withdrawals. Trials were rated for methodological quality in duplicate using the Jadad scale and scored out of a maximum of five (19, 20). An inter-rater agreement was also calculated using the κ statistic (21).
Five outcomes were analyzed: total symptom score, total nasal symptom score, nasal eosinophils, nasal airflow, and nasal IL-4. Variables such as allergen dose level, break reaction time, driving lateral position, histamine skin test, metacholine challenge test, nasal congestion symptom score, nasal peak inspiratory flow, secretion weight, peak expiratory flow, performance test, quality of life, total asthma symptom score, rescue medication use, sleeping score, tracking error, vigilance, and products from nasal lavage (albumin, eotaxin, IL-5, IL-8, neutrophils, and tryptase) were evaluated in some studies; these were not included in this analysis because of a lack of a sufficiently sized patient population or as the relevance of these measures was deemed insufficiently robust.
Two independent reviewers extracted data from the selected articles, reconciling differences by consensus. It was planned to perform an intention-to-treat analysis and to attempt to include dropouts in the analysis, if Last Observation Carried Forward for all continuous scores were available; if not, we included in the analysis patients with a final assessment. When the results were only presented in graphs, these were digitalized and then converted to numbers using DigitizeIt 1.5.7 (DigitizeIt 2003; Bormann, Braunschweig, Germany). Additionally, wherever possible we contacted investigators to obtain more information for data extraction.
The outcome data analyzed was quantitative continuous in nature (symptom scores, airflow measurements, cell counts, and cytokine levels). In the original studies, some scoring systems and scales were used for symptoms (most frequently a daily assessment of symptoms entered on a patient's diary and subsequently summarized and averaged), while nasal airflow was measured by active anterior rhinomanometry and was reported as the sum of recorded airflow through right and left nostrils during a period of time; three or more airflow measurements were performed for each patient and the mean was recorded when reproducible values were achieved. Interleukin-4 levels were measured using an ELISA after specimens were obtained by nasal lavage and nasal cytology was performed after nasal lavage fluid centrifugation after which the pellet was transferred to slides and finally stained for cell differential counts by light microscopy.
For the analysis of the treatment effect from baseline, the investigators of each trial provided the standard deviation (SD) of the distribution separately for baseline and post-treatment values for both the active treatment and control groups. Post-treatment SDs were then combined for the analysis, assuming independence. When the outcome variables continuous data were expressed in different scales, we estimated the standardized mean difference (SMD); however, when a common scale was used, we calculated the weighted SMD. The SMD expresses the difference in mean values between drug and placebo in units of SDs and it has the important property that its value does not depend on the measurement scale. Heterogeneity was calculated with the chi-square and the I-square (I2) tests. The I2 statistic explores the degree of heterogeneity, and can be interpreted as the proportion of the observed discrepancy in the estimation of the effect, within a group of trials, which cannot be accounted for by random variation. Substantial heterogeneity exists when I2 exceeds 50% (22, 23). All results are reported with 95% confidence intervals (95% CI) and all P-values are two-tailed.
Given the significant heterogeneity found among the results of the included studies, the random-effects model was used in all the comparisons, because it accounts for the possibility of significant inter-study heterogeneity and provides wider CIs than fixed-effect models (24). Analysis was performed using RevMan 4.2.8 program (RevMan, 2005. The Cochrane Collaboration, Oxford, UK).
The primary search identified 229 articles, 57 of which were potentially relevant trials on desloratadine in AR (Fig. 1). A total of 32 studies were randomized and 13 publications met inclusion criteria for the meta-analysis(25–37). The κ statistic for inter-rater agreement on study eligibility was 0.80 (95% CI, 0.75–0.95). A group of randomized trials were excluded from the review for the following reasons: 12 were designed to study outcomes other than the aims of this review, 3 lacked suitable data, 2 compared desloratadine with compounds not valid for this meta-analysis, 2 were pharmacokinetic evaluations, and 1 was designed for safety.
Table 1 outlines the characteristics of the studies and subjects included in the meta-analysis. The trials included 3108 patients who completed treatment and data analyzed. The median age of the participants was 32 years (range: 12–65 years). Each trial included a median of 45 participants (range: 16–1179). The median duration of treatment was 2 weeks. Three studies treated patients with single doses, two for 1 week, four for 2 weeks, and four for up to 4 weeks.
Table 1. Characteristics of trials and participating subjects
Age (mean), years
DL dose, mg/day
Patients per group
(S), spring season; (F), fall season; SD, single dose; FEX, fexofenadine; LCZ, levocetirizine; DL, desloratadine; RCT DB CX: randomized clinical trial, double-blind, crossover.
While 13 studies met inclusion criteria, we identified 20 comparison groups, as the study investigators evaluated placebo, levocetirizine (n = 5) and fexofenadine (n = 2) as control groups. Total symptom score, total nasal symptom score, and nasal airflow included postexperimental challenge evaluations.
Data on total symptom score were obtained from 11 treatment comparisons (n = 3108); 7 treatment comparisons were available for the total nasal symptom score (n = 2883), 5 for nasal peak inspiratory flow (n = 292), 5 for nasal airflow (n = 438), 4 for nasal eosinophils (n = 133) and 3 for IL-4 (n = 88).
Methodological quality of included studies
All of the included trials were randomized, double blind, and controlled. All studies secured informed consent. Each trial reported dropouts and withdrawals and analyzed patients completing the trial; dropout rate varied between 0% and 14%. Based on Jadad's criteria, 10 studies received a 5/5 quality score, and 3 studies were rated 4/5. The κ score for inter-rater agreement on methodological quality was 0.90 (95% CI: 0.80–1.0).
Drug comparison analysis
A significant inter-study heterogeneity was found in all comparisons except for nasal airflow. Table 2 shows the results of each drug comparison.
Table 2. Results of drug outcome comparisons for studies included in the meta-analysis
I 2 (%)
DL, desloratadine; LCZ, levocetirizine; I 2, I-square statistic.
Total symptom score
−2.75 to −0.51.
−0.20 to 1.27
Total nasal symptom score
−0.91 to −0.42
0.10 to 0.55
−2.57 to 0.01
−0.74 to 3.49
−5.01 to 0.64
−0.91 to −0.42
The analysis of total symptoms scores included 3108 subjects, 1553 of whom received desloratadine and 1555 placebo (Fig. 2). In an analysis of patients not studied in a challenge setting (desloratadine n = 1463; placebo n = 1465), desloratadine was associated with a significant reduction in total symptoms scores when compared with placebo (SMD −1.79; 95% CI −3.10 to −0.47; P = 0.008). Similarly, the desloratadine-to-placebo comparison for total nasal symptoms score (n = 2883 subjects) demonstrated a significant reduction favoring desloratadine (SMD −0.66; 95% CI: −0.91 to −0.42; P < 0.001). Comparison with levocetirizine as an active comparator in terms of total nasal symptoms was possible in a very small population (n = 86), which showed no measurable difference between the two compounds.
Objective measures of nasal obstruction
A total of 438 subjects from seven studies formed the population for the comparison between desloratadine (n = 218) and placebo (n = 220) in terms of nasal airflow (both following clinical treatment and in the allergen challenge setting). The difference between desloratadine and placebo significantly favored desloratadine [SMD 0.32; 95% CI: 0.10 to 0.55; P = 0.005 (Fig. 3)].
For the comparison between desloratadine and placebo in terms of nasal eosinophils, four studies with 133 subjects were included. This comparison demonstrated an effect that significantly favored desloratadine over placebo (SMD −1.28; 95% CI −2.57 to 0.01; P = 0.05). The corresponding comparison between desloratadine and placebo in terms of IL-4 from nasal lavage included three studies with 88 subjects; there was no demonstrable difference between the two groups. Similarly, comparisons between desloratadine and levocetirizine did not show a difference favoring one over the other as measured by nasal eosinophils or nasal lavage IL-4.
This meta-analysis of the literature assessed the efficacy of desloratadine in the treatment of AR, identifying a total of 13 randomized, controlled trials that met the study's inclusion criteria. We confirm that the effects of desloratadine seen in individual studies of heterogeneous design and duration are borne out following a systematic review. Indeed, this study demonstrates for the first time that the efficacy of desloratadine vs placebo in AR is supported by Ia evidence, the highest ranking in evidence-based medicine. Desloratadine had an overall beneficial effect on the symptoms of AR, as measured by total symptoms scores and by total nasal symptoms scores, although there was heterogeneity among the individual studies. As such, this meta-analysis ties together data from disparate published clinical trials in AR involving patients with perennial or seasonal AR, who were treated at the clinical dose over a variety of follow-up periods. This meta-analytic approach has permitted us to analyze together data on AR symptoms, which have been gathered from short-term challenge studies and longer clinical trials; the overall results of the meta-analysis demonstrate the consistency of desloratadine across the published body of literature in AR.
As the symptom nasal congestion is a component score of both total symptoms and total nasal symptoms scores, we wanted to assess whether there was a consistent effect of desloratadine on more independent measures of nasal blockage. For this reason, the meta-analysis included studies with objectively measured endpoints of nasal obstruction. Overall, the study confirmed that desloratadine is associated with a homogeneous effect on nasal airflow in the included studies. This provision of type Ia evidence supporting the efficacy of desloratadine in terms of improving nasal airflow more than placebo serves to confirm that nasal symptomatic benefits are truly accompanied by an objective decrease in nasal obstruction. The mechanism by which antihistamines mediate their clinical effects has been the subject of many studies in recent years. While undoubtedly histamine H1-receptor antagonism to block the direct effects of histamine derived from mast cells is central, other effects have been noted (38). These anti-inflammatory effects, usually noted in the in vitro setting, relate to the modulation of cytokines, adhesion molecules, and effector cells, such as eosinophils. While studies have demonstrated anti-inflammatory effects of desloratadine at low concentrations in vitro (39), in this meta-analysis, we examined the effects on cells and mediators in patients with AR following oral dosing. The analysis noted a statistically significant effect of desloratadine on nasal eosinophils, which indicates that oral dosing with desloratadine can modulate allergic inflammation in a relevant fashion.
Direct comparisons among second-generation antihistamines are relatively infrequent in the literature, when compared with placebo-controlled comparisons. In the current meta-analysis, we were able to include data from multiple comparative trials; however, they were uniformly small in size, thus decreasing the robustness of results obtained. Overall, there were no results significantly favoring levocetirizine or fexofenadine over desloratadine in terms of AR symptoms scores, nasal airflow measurements or allergic inflammatory cells and mediators. This helps to confirm that small or incremental differences in symptom control between antihistamines in individual studies are not borne out by meta-analytic techniques. The difference between the current meta-analysis and other systematic reviews of antihistamines is that we applied validated methodologies for the selection and analysis of studies; less complete search strategies and data pools permit only selective interpretations of findings that fall short of type Ia evidence.
In conclusion, this is the first meta-analysis of clinical data regarding an antihistamine to show significant beneficial effects on not only symptoms of AR, but also objective measures of nasal blockage and to some extent, allergic inflammation. This type Ia evidence augments the existing profile of desloratadine in AR obtained from individual controlled clinical trials.
This study has been partially supported by ARMIA (Associazione per la Ricerca delle Malattie Immunologiche e Allergiche – Genova) and GA2LEN (Global Allergy and Asthma European Network). We thank the unrestricted collaboration of Profs William Berger, Michael M. Cyr, Adrian Daly, Cédric Deruaz, Friedrich Horak, Daniel K. C. Lee, Richard Lorber, Eli O. Meltzer, Giovanni Passalacqua, and Estelle Simons.
Dr Giorgio Walter Canonica has received research grants and lecture fees from Alk-Abello, Anallergo, Allergy Therapeutics, A. Menarini, Almirall, Altana, AstraZeneca, Boehringer Ingelheim, Chiesi Farmaceutici, Genentech, Gentili, Glaxo Smith Kline, Lofarma, Merck Sharp & Dome, Novartis, Pfizer, Schering Plough, Stallergenes and UCB Pharma. Dr Martin Penagos has received research grants from Aventis and Glaxo Smith Kline. Dr Enrico Compalati and Dr Francesco Tarantini have not declared conflicts of interest.