A Multicenter Trial of Specific Local Nasal Immunotherapy
Abstract
Objective: To assess the efficacy and safety of specific local nasal immunotherapy (LNIT) in powder form in patients with allergic rhinitis, using subjective and objective parameters.
Study Design: A double-blind randomized multicenter trial of 102 patients with allergic rhinitis who were treated with specific LNIT for 8 consecutive months.
Methods: After identifying allergens with the skin prick test and sensitization threshold dose with the specific nasal provocation test, 102 patients were selected, of whom 55 were allergic to mites and 47 were allergic to Graminaceae or Parietaria pollen. The specific treatments were self-administered using an insufflator in two phases (phase 1: increasing doses; phase: 2, maintenance dose). Patients were evaluated before and after 32 weeks of treatment by subjective analysis of their self-reported symptoms and by objective analysis of nasal provocation test, nasal resistance by anterior rhinomanometry, and mucociliary clearance time.
Results: Clinical efficacy of LNIT for allergy to mites and pollens was confirmed by the differences in the symptoms score between the active group and the placebo group. The nasal provocation test results confirmed that this difference was statistically significant. The rhinomanometric analysis gave positive results for the treated group mainly in LNIT for mites. No differences in mucociliary clearance time were found.
Conclusions: Specific LNIT is effective for allergic rhinitis and appears to offer considerable advantages over other hyposensitization methods. It can be done at home, patient compliance is good, and the treatment is safe.
INTRODUCTION
For years the manifestations of atopy, including allergic rhinopathy, have been considered systemic disorders, to be generally treated with allergen-specific immunotherapy. This approach normally involves a weekly subcutaneous injection of an extract of the allergen, in solution, in increasing doses until a standard maintenance dose is reached. This dose is then injected subcutaneously on a regular basis-at intervals of approximately 20 days-normally for not less than 3 years for perennial allergens.
The aim of this therapy is to induce a gradual change in the differentiation of T lymphocytes so that they preferentially become T-helper 1 effector cells, which produce interleukin-2 and interferon gamma, rather than T-helper type 2 cells, which mainly synthesize interleukin-4. These changes in the cytokine profile lead to a gradual reduction in immunoglobulin E synthesis and inhibition of mast cell activation, thus impeding the release of the chemical mediators of immune inflammation. The resulting local physiopathological changes affect the whole progression of the disorder and halt the production of symptoms of rhinitis. The utility of allergen-specific immunotherapy in treating nasal allergies has been confirmed in double-blind clinical trials, reported in a World Health Organization Position Paper.1
In recent years local therapies have been aimed at overcoming the clinical manifestations of allergy in single organs. For allergic rhinitis, treatments acting on the symptoms include local steroids, antihistamines, cromolyn, anticholinergic agents, and so forth. Therapies directed against the causes of allergy require specific desensitization for each patient, administered directly into the nose. The allergen extract, in powder form (Allerkin, Lofarma, Milan, Italy), is given using a nasal insufflator in a two-phase schedule. In phase 1 the doses gradually increase up to the maximum tolerated dose and in phase 2 the maximum tolerated dose is maintained with once-weekly inhalations for 1 or 2 years at least (advisable for perennial allergens such as mites and Parietaria in Southern Italy).
Specific nasal immunotherapy thus applies the same principles as conventional immunotherapy while exploiting recent knowledge to permit local treatment in individuals with limited allergic reactions. Various studies have documented the clinical efficacy of this approach, using aqueous extracts2-6 or extracts in powder form,7-22 and rating patients' subjective impressions using clinical scores and objective variables such as specific nasal provocation tests, mucociliary clearance time, and levels of intercellular adhesion molecule-1. The World Health Organization now lists nasal immunotherapy, together with oral and sublingual approaches, as alternative therapeutic options to the subcutaneous route.1 This preliminary trial was designed to obtain further confirmation of the validity of local specific nasal immunotherapy (LNIT) in patients with allergic rhinitis monosensitive to perennial allergens such as household dust mites (Dermatophagoides) or seasonal allergens such as the pollen from Graminaceae or Parietaria.
MATERIALS AND METHODS
Patients
This double-blind trial comprised 102 patients (54 females, 48 males; aged 9-49 y; mean age, 26 y) with nasal hyper-reactivity symptoms (rhinorrhea, sneezing, and obstruction). They were assigned to receive either specific LNIT (Allerkin, Lofarma, Milan, Italy) or placebo. Patients were selected between 1991 and 1993 from among those attending the ENT departments at the seven centers participating in this trial (“Federico II” University, Naples; University of Turin; University of l'Aquila; University of Milan; S. Martino Hospital, Genoa; Policlinico Gemelli, Rome; and La Sapienza University, Rome).
Each patient was examined to 1) establish a description of clinical signs and symptoms; 2) identify causal factors; and 3) assess morphological changes by anterior rhinoscopy. To identify the allergens causing the rhinitis, three tests were performed. First, the skin prick test was used to identify the inhaled allergens (pollens and mites), which are most widely responsible for nasal allergies. Second, specific nasal provocation tests were performed for pollens and mites, using active anterior rhinomanometry, as follows: baseline assessment to establish which nasal fossa had least resistance; placebo challenge (lactose powder, Allerkin test); rhinomanometric assessment; and allergen challenge with doses of 2.5, 5, 10, 20, 40, 60, 80, 120, 160, and 240 allergen units (AU) (Allerkin test) at 10-minute intervals, recording the rhinomanometric tracing at the end of each challenge. The test was considered positive for the allergen dose that caused an increase in resistance of 100% or more. Finally, nasal microbiological tests were done to identify any concomitant infectious pathogens such as bacteria, mycetes, Chlamydia, and so forth.
Inclusion and Exclusion Criteria
Patients had to be monosensitized to Dermatophagoides or grass or Parietaria pollens, have a history of allergy to one of the allergens considered, and have a positive skin prick test and a nasal provocation test for one of these allergens.
Patients were ineligible if they had significant structural abnormalities of the nasal cavities (deviation of the septum, polyposis of the turbinates), neoformations such as polyps and tumors, concomitant infectious agents (bacteria, mycetes, or Chlamydia), or sensitization to more than one allergen.
Patients fulfilling these criteria were admitted for specific LNIT for the allergen causing their rhinitis. Of the 102 patients selected, 55 (54%) were positive for mites and 47 (46%) were positive for Parietaria or Graminaceae.
Using a randomization list (Table I) patients were assigned to the active group (54 patients), which received specific LNIT according to the schedule described below, or to placebo (48 patients), in which case they followed the same schedule but instead of the allergen extract they were given lactose powder formulated so it was indistinguishable from the active powder. Each active product contained a single allergenic extract: from mites (50% Dermatophagoides pteronyssinus and 50% Dermatophagoides farinae); from mixed Graminaceae (Phleum pratense, Poa pratensis, Dactylis glomerata, Festuca elatior, Lolium italicum); and from Parietaria mix adsorbed on lactose (inert excipient) and dehydrated.
|
The extracts were titrated in allergenic units, 1 AU being one fortieth of the mean provocative threshold in a nasal provocation test done in a group of allergic volunteers.18 The extracts were standardized by radioallergosorbent test inhibition against an internal reference standard.
Dosage Schedule
In phase 1 the doses increased over time. The allergenic extracts were given daily into alternate nostrils, at doses of 2.5, 5, 10, 20, 40, 60, 80, 120, 160, and 240 AU. Each dose was repeated six times until the highest dose had been given.
Phase 2 was the maintenance phase. Once patients had reached the highest dose-240 AU-they took it weekly, alternating the nostrils, for 1 year. This regimen applied for all three allergens employed in this trial.
Clinical and Instrumental Investigations
Before treatment started and after 32 weeks of treatment, subjective and objective investigations were undertaken to quantify symptoms and clinical findings. For the subjective analysis, patients rated the symptoms of obstruction, rhinorrhea, and sneezing on a scale of 0 to 4 (0 = no symptom; 1= mild; 2 = moderate; 3 = serious; 4 = unbearable). At the end of the trial the scores for each symptom became the basis for classifying each patient as either “cured”(symptom present at baseline but no longer at the end of the trial); “improved”(score for symptoms at baseline higher than at the end of the trial); “unchanged”(same scores at baseline and end of trial); or “worsened”(end-of-trial scores higher than baseline).
For the objective analysis, the clinical picture was analyzed on the basis of the following instrumental findings. First, nasal reactivity threshold, as determined by rhinomanometry in the specific nasal provocation test with the allergen.
Second, nasal resistance, as measured by static and dynamic rhinomanometry, following the methods routinely employed at Naples University. A computerized rhinomanometer was used (Rhinospir-164, Markos, Monza, Italy) with a mask. Nasal resistance was measured at a preset pressure of 150 Pa, in accordance with the recommendations of the International Standardization Committee.24 The patient was examined in a seated position (static rhinomanometry); after he or she had been supine for 30 minutes (dynamic rhinomanometry, to detect changes in nasal resistance due to the position, with increased blood flow to the inferior turbinates); and after local vasoconstrictors were applied, to check whether any increase in nasal resistance was caused by vasomotor phenomena or obstruction due to other pathologies such as deviation of the septum, nasal or rhinopharyngeal masses, and so forth. Phenylephrine HCl was sprayed into each nostril 10 minutes before measurements were to be made, and again after about 5 minutes (functional rhinomanometry).
Third, mucociliary clearance time was measured by applying vegetable charcoal mixed with 3% saccharine (approximately 5 mg) onto the mucosa at the medial face of the inferior turbinate, about 1 cm from its head. Patients were asked to breathe regularly, without blowing their noses, and to report when they felt the sweet taste in the pharynx. An interval of 10 to 20 minutes was considered normal.
For each of these investigations, we used a rating system taking account of the normal reference values:
Nasal provocation test: threshold dose > 60 AU = test negative; threshold dose 60 AU = 1; threshold dose 40 AU = 2; threshold dose ≤ 20 AU = 3.
Rhinomanometry: <0.25 Pa/s or cc/s = 0 (normal reference value); 0.25-0.27 Pa/s or cc/s = 1; 0.27-0.3 1.2 Pa/s or cc/s = 2; >0.3 Pa/s or cc/s = 3.
Mucociliary clearance time: <20 minutes = 0 (normal reference value); 20 to 25 minutes = 1; 26 to 30 minutes = 2; >30 minutes = 3.
Statistical Analysis
Results were analyzed statistically using the Markov Chain method to assess changes or persistence of subjective symptoms and instrumental findings at baseline and the end of treatment (32 weeks). Differences between these times were assessed using the χ2 test or Fisher's test.
In patients allergic to mites, we checked the probability of treatment being successful in the two groups using logistic regression, and the likelihood ratio test. The odds ratio calculated from the logistic regression was used to assess the probability of being “cured” or “not cured” in the active treatment and placebo groups.
In patients allergic to pollens we could not make any further statistical analysis because the case list did not meet the necessary theoretical conditions (the number of patients was insufficient).
The same procedures were used to assess a more cautious hypothesis that considered patients “cured” only if all symptoms and scores present at baseline had disappeared totally at the end of the trial. This was done in the larger group, i.e., in those patients allergic to mites.
RESULTS
Symptoms
The clinical picture of allergic rhinitis comprises a few typical symptoms: nasal respiratory obstruction, rhinorrhea, and sneezing. Patients reported these symptoms with varying intensity and frequency. Tables II and III set out the mean scores before and after treatment, with statistical analysis of the subjective symptoms. Results are as follows.
|
|
Patients allergic to mites (Table II). Comparison of the active treatment and placebo groups revealed the following that there was a reduction in the mean score for nasal obstruction after treatment, and compared with placebo, there was a significantly higher percentage (P < .0001) of patients in the LNIT group who showed improvement (82.8% vs. 11.5%) or were cured (6.9% vs. 0%).
Logistic regression applied to the end-of-trial results showed that the probability of a positive response to LNIT was 89.6% compared with 11.5% for placebo. The difference was significant (P = .0000). The odds ratio of cured to noncured patients was 9:1 in favor of the active treatment, meaning that for every nine patients cured or showing improvement with the active treatment, only one responded similarly to placebo.
Further, the percentage of patients with rhinorrhea who improved or were cured was much higher in the active treatment group than the group with placebo (62.5% vs. 14.3% and 18.7% vs. 0%; P = .0007). Logistic regression showed that the probabilities of a response were respectively 81.2% and 14.3% (P = .0001) for the two groups, and the ratio of cured to noncured patients was 7:1.
The percentage of patients reporting sneezing improved with LNIT somewhat more than with placebo (46.7% vs. 23.1%) and the gap between those cured and those not (26.7% vs. 0%) increased in favor of LNIT (P = .0046). The probabilities of improvement or cure were respectively 73.3% and 23.1% for LNIT and placebo, the difference being highly significant (P = .0066). The odds ratio was 4:1 in favor of LNIT.
Patients allergic to pollens (Table III). Among the patients with nasal obstruction, 17 responded well to treatment (cured or improved), and 6 remained unchanged or worsened. The χ2 test comparing these two groups showed there were more responders (73.9%) among patients assigned to LNIT than among those given placebo (40%) and the difference was significant (P = .0366).
Twelve patients with rhinorrhea (75%) responded to LNIT with either cure or improvement. The χ2 test showed this proportion was significantly higher than in the placebo group (2 patients, 20%; P = .0062). Eight patients reporting sneezing at the start of the trial responded to active therapy (57.1%, cure or improvement), whereas only one patient in the placebo group (11.1%) showed improvement. The difference was significant (P = .027).
Instrumental Examinations
Tables IV, V, and VI set out the findings of rhinomanometry, mucociliary clearance time, and nasal provocation test. Statistical analysis produced the following picture.
|
|
|
Patients allergic to mites (Table VII). Rhinomanometry showed that responses in the group given active LNIT were better than with placebo (10.3% vs. 3.8% improvements and 27.6% versus 3.8% cures). Two-way logistic regression indicated that the probability of improvement or cure with LNIT was 37.9% compared with 7.7% for placebo. The odds ratio for patients cured or improved with LNIT compared with placebo was 6:1.
|
Mucociliary clearance time returned to normal in 8% of patients given specific LNIT. None of those given placebo showed any improvement.
The nasal provocation test in patients given LNIT showed a lowering of the threshold for reactivity in a significantly higher proportion of cases than controls. Improvement or cure was reported in 31% and 48.3% of the patients who had undergone LNIT, compared with 7.7% and 15.4% of those given placebo. The differences were significant (P < .0001). Two-way logistic analysis of the end-of-trial findings showed that the response to LNIT was 79.3%, much higher than with placebo (23.1%). The odds ratio was 4:1, indicating that for every four patients cured or improved with therapy, only one obtained a similar result with placebo.
Further analysis in the group allergic only to mites showed that among those whose score was above zero at the start of the trial but whose problems had disappeared after treatment, there was no difference between “cures” and “improvements,” in that they amounted to 20.7% in the LNIT group compared with 0 in placebo. This difference was significant (P = .0243, Fisher's test). The calculation only refers to rhinomanometry, nasal provocation test, and mucociliary clearance time.
Patients allergic to pollens (Table VIII). Rhinomanometry showed that 12 patients were cured or improved (50%). Among the patients showing no change, 72.7% had had no symptoms at baseline and 27.3% had reported only mild symptoms. One patient's symptoms worsened. Although there was a larger proportion of responders to LNIT, the difference from the placebo group was at the limits of significance. Mucociliary clearance time findings did not reach statistical significance (P = .6671). In the LNIT group the nasal provocation test showed 13 patients responding to therapy and 11 with no change. One case worsened. The χ2 test, used to compare responders (cured and improved) and nonresponders (no change or worsened) showed that significantly more patients responded to LNIT (52%) than to placebo (22.7%) (P = .0394).
|
DISCUSSION
Local immunotherapy for allergic rhinitis showed significant efficacy in this trial. The results were better in the allergies brought on by perennial allergens than in the seasonal allergies due to pollens. Active treatment was undoubtedly effective and was always superior to placebo in regard to the proportion of patients cured or improved, and the probability of a response to treatment.
We investigated several aspects including 1) changes in nasal mucosal reactivity to the allergen using nasal provocation test; 2) objective symptoms studied by rhinomanometry and a mucociliary clearance test; and 3) subjective symptoms (nasal obstruction, sneezing, and rhinorrhea).
All patients with perennial allergy to mites had symptoms at the start of the trial, and the high percentage of clinical improvement-documented by the responses to nasal provocation test-corresponded to equally significant improvements in objective and subjective symptoms.
In patients with seasonal allergies to Graminaceae and Parietaria, the percentage of clinical improvement did not always correspond to similar relief of subjective and objective symptoms. This depends on what stage the allergy was at when the trial started. In these cases baseline tests were not done when the allergy was in an acute phase. This explains the low proportions of patients with severe rhinorrhea and sneezing at the beginning of the trial (55% and 50%). As a consequence, mucociliary clearance time was usually only partly impaired in a small proportion of cases (10%-27%). These subjective (rhinorrhea and sneezing) and objective (rhinomanometry) symptoms could therefore not be expected to show major changes.
Nasal obstruction, on the other hand, is a basic feature of the clinical picture in patients with nasal atopy, although it is naturally worse in the acute phases. Relief of nasal hyper-reactivity with LNIT improves this situation, as confirmed by the patients' reports and the rhinomanometric findings.
CONCLUSION
Allergic rhinitis still poses problems today and must be studied not only as the clinical manifestation of a systemic disorder but also, perhaps primarily, as a pathology specifically affecting the nose. The symptomatic picture of allergic rhinitis reflects the morphological and functional features of the nasal cavities and the reactions of their mucous membranes. These features must be borne in mind when selecting the best therapeutic approach, including topical drugs, for each patient. Morphological anomalies in the nose must therefore be taken into account too. The same clinical signs are often caused by different morphological features, and objective examination does not always indicate the extent of vasomotor involvement. Static, dynamic, and functional rhinomanometry is therefore essential, as recommended at Naples University.
It is equally important to identify the causes of allergic rhinitis with a view to selecting appropriate therapy. Nasal provocation tests are more reliable than systemic tests, although these are useful for first-level screening in patients with rhinitis presumably caused by allergies. They demonstrate the causal relationship between local allergic reactions in the nasal mucosa and the allergen responsible. They identify local nasal symptoms in patients with no systemic alterations, and help establish the main allergen causing symptoms in patients with multiple sensitizations.
Allergic symptoms can be treated by specific LNIT using the responsible allergen in two stages as shown in this trial.
Although the results reported here refer only to 8 months' therapy, the group treated with LNIT showed considerably enhanced reactivity at the end of the dose-increasing stage (phase 1), and clinical signs related to the nasal allergy were much reduced by the end of the trial. Clearly, longer follow-up is needed to see how long maintenance treatment needs to continue and how best it should be scheduled.
Specific LNIT appears to offer considerable advantages over other hyposensitization methods. The proportion of failures is limited and might be further reduced once more appropriate and “personalized” schedules are established. Because it can be done at home, patient compliance is better. Complications are very few: no severe systemic reactions have been observed, and only mild local reactions, such as light sneezing, have been reported. This seems to be in contrast with many studies performed with aqueous extracts, in which local side effects were frequent and troublesome. The explanation is in the pharmaceutical formulation in dry powder, which is much better tolerated, as has been shown in many previous clinical studies (reviewed by Passalacqua et al.23). Naturally, the success of this therapeutic approach relies on correct clinical investigations and reliable diagnostic methods, especially as regards factors that may interfere with the pathogenesis of the clinical manifestations of allergic rhinitis.
ACKNOWLEDGMENT
We thank Prof. M. Recchia (Statistical Department, University of Milan) for the helpful assistance in statistical analysis, and for their collaboration in the collection of clinical data, we thank F. Beatrice, E. Siccardi (ENT Department, University of Turin); I. Costa, G. Massari (ENT Department, University of Milan); L. Bellussi, C. Biagini (ENT Department, University of Siena); A. Canessa (ENT Division, S. Martino Hospital, Genoa); F. Bellioni (ENT Department, Policlinico Gemelli, Rome); F. Filiaci and A. Ciofalo (ENT Department, La Sapienza University, Rome).
