The Laryngoscope C 2014 The American Laryngological, V
Rhinological and Otological Society, Inc.
Controlled Trial for Long-Term Low-Dose Erythromycin After Sinus Surgery for Chronic Rhinosinusitis Boris R. Haxel, MD, PhD; Meike Clemens; Niki Karaiskaki, MD; Uta Dippold, MD; Lisanne Kettern, MD; Wolf J. Mann, MD, PhD, FACS Objectives/Hypothesis: The efficacy of macrolides in chronic rhinosinusitis (CRS) is still under controversy. To date, only two double-blind, placebo-controlled studies have been published with differing results. None of these studies investigated the possible benefit of macrolides in the postoperative period. We conducted an investigator-initiated clinical trial using 250-mg erythromycin once a day over a period of 3 months, beginning the administration of either erythromycin or placebo 2 weeks after a surgical intervention for CRS. Study Design: Randomized double-blind, placebo-controlled trial. Methods: The concentrations of eosinophilic cationic protein (ECP) and myeloperoxidase in nasal secretion were chosen as primary outcome measures. Additionally, as a secondary outcome measure, changes in the Sino-Nasal Outcome Test-20 score, olfaction, saccharin transit time, nasal endoscopy score, and self-rating of nasal health using a visual analogue scale were evaluated. Results: Sixty-seven patients after surgery for CRS with or without nasal polyps were screened, and 58 patients were randomized to the study groups. For the primary outcomes, the concentrations of ECP changed from 176.4 ml/l 6 79.0 to 226.1 ml/l 6 200.6 in the erythromycin group and from 186.9 ml/l 6 36.0 to 192.9 ml/l 6 189.2 in the placebo group; no statistical differences were found. Of the secondary outcomes, only the nasal endoscopy score showed a statistically significant improvement in the erythromycin group (from 2.6 6 1.4 to 1.9 6 1.5 points) compared to the placebo group (from 2.5 6 1.3 to 2.6 6 1.5 points). The subgroup of patients without nasal polyps in the erythromycin group showed a tendency to improvement in some secondary outcome criteria. Conclusions: A general recommendation for long-term, low-dose erythromycin treatment after surgery for CRS cannot be given. In patients with CRS without nasal polyps, a tendency to improved parameters was detected. Key Words: Macrolides, sinus surgery, CRS, evidence-based medicine. Level of Evidence: Ib. Laryngoscope, 00:000–000, 2014
INTRODUCTION Rhinosinusitis is regarded as a multifactorial disease that causes different symptoms such as nasal obstruction, headache, nasal discharge, and olfactory dysfunction. Chronic rhinosinusitis (CRS) is defined as a state of inflammatory disease that occurs for longer than 12 weeks and is diagnosed by nasal endoscopy and computed tomography (CT) scan of the sinuses. CRS can
Additional Supporting Information may be found in the online version of this article. From the Department of Otorhinolaryngology, Head and Neck Surgery, AMEOS Klinikum Haldensleben (B.R.H.), Haldensleben; and the Department of Otorhinolaryngology, Head and Neck Surgery, University Medical Center of the Johannes Gutenberg University (B.R.H., M.C., N.K., U.D., L.K., W.J.M.), Mainz, Germany. Editor’s Note: This Manuscript was accepted for publication October 29, 2014. Presented at the 84th Annual Meeting of the German Society of Oto-Rhino-Laryngology, Head and Neck Surgery, Nuremberg, Germany, 2013. The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Boris R. Haxel, MD, PhD, Department of Otorhinolaryngology, Head and Neck Surgery, AMEOS Klinikum Haldensleben, Kiefholzstr 27, D-39340 Haldensleben, Germany. E-mail: [email protected] DOI: 10.1002/lary.25052
Laryngoscope 00: Month 2014
present as cases with nasal polyps (CRSwNP) and without nasal polyps (CRSsNP), which seem to have a different underlying pathophysiology.1 Possible causes for the inflammation in the nasal mucosa include staphylococcal superantigen, aspirin-exacerbated respiratory disease, biofilm, or fungi, just to mention the most common.2–4 Present and future research efforts aim to find differences in individual phenotypes and endotypes such as cytokine levels in nasal secretion to be able to distinguish between different specifications of the underlying inflammatory disease.5,6 Therapeutic options are mainly topical nasal steroids, antibiotics in acute exacerbations, and short-term oral steroids. In patients who do not improve after a conservative treatment, a surgical approach (functional endonasal sinus surgery) is recommended.1 Surgery is effective to control local inflammation and improves the quality of life7 but is usually followed by long-term topical steroid treatment. Especially in CRSwNP patients, the risk of recurrent disease is higher than in CRSsNP patients, reaching rates of about 20%.8 In some studies, long-term, low-dose macrolides are proposed for medical treatment of CRS.9,10 Here, macrolides are thought to have antiinflammatory or immunomodulatory capacities besides their antimicrobial effects. Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
1
This is based on the observation of blockage of the production of interleukin-8 (IL-8) and tumor necrosis factor, the effects on neutrophil migration and adhesion, and modulation of synthesis and secretion of mucus.11 In clinical trials of CRS, macrolides showed improvements in symptoms and endoscopic findings, decrease in polyp size, decrease of radiologically conspicuous mucosal changes in the sinuses on CT scan, and reduction in neutrophils and interleukin (IL)28 levels in nasal discharge—as well as an improvement in saccharine transit time.10,12 Although these results showed promising effects of macrolides in CRS, most of these studies were conducted as open and noncontrolled clinical trials. To date, only two double-blind, placebo-controlled (DBPC) studies have been published with differing results.13,14 In the Wallwork study, CRS patients without topical or systemic treatment 4 weeks prior to the start of the macrolide therapy were included; whereas in the Videler study, patients who suffered from recalcitrant CRS despite previous surgery and intensive medical treatment were evaluated.13,14 None of these studies investigated the possible benefit of macrolides in the immediate postoperative period. In the 2012 European Position Paper on Rhinosinusitis (EPOS) guidelines, macrolides for treatment of CRS is recommended for patients with CRSsNP (level of evidence Ib, strength of recommendation C).1 We conducted a randomized, double-blind, placebocontrolled clinical trial with low-dose erythromycin over a period of 3 months directly after a surgical intervention for CRS. The changes in concentrations of eosinophilic cationic protein (ECP) and myeloperoxidase (MPO) in nasal secretion were chosen as primary outcome measures. Additionally, we followed clinically and functional parameters and possible side effects of the study medication.
Gutenberg University, Mainz, Germany, as an investigatorinitiated study.
Study Medication As study medication, 250-mg erythromycin once a day or placebo was administered over a period of 3 months, beginning 2 weeks after the surgical intervention for CRS. A randomization list was generated by a computer program in the Interdisciplinary Center for Clinical Trials and was confidentially stored in this institution. A copy of this list was only forwarded to the pharmacy of the university hospital. The pharmacy provided the packing and blinding of the study medication. The erythromycin 250-mg tablets were packed into capsules. Identical empty capsules served as placebo.
Concomitant Medication All patients used nasal irrigation with saline at least twice a day and were treated with a topical nasal steroid (fluticasone furoate) once a day (27.5 mg), beginning after the surgical intervention and continuing for the entire study period.
Primary and Secondary Outcome Parameters The concentrations of ECP and MPO in the nasal secretions were chosen as primary outcomes. Additionally, changes in the nasal endoscopy score, saccharin transit time, olfaction, Sino-Nasal Outcome Test (SNOT-20) score, and a subjective nasal health status determined by a visual analogue scale were evaluated. Additionally, the preoperative CT scans of the paranasal sinuses were analyzed and the Lund-Mackay score was noted, as well as the status of atopy by using a skin prick test and in vitro methods (specific immunoglobulin E [IgE] for the most common environmental allergens in the ImmunoCAP testing). Furthermore, the blood IgE-levels of the patients were determined, and it was noted whether the patient suffered from CRSsNP or CRSwNP.
Follow-up MATERIALS AND METHODS Patients In a prospective study, patients of both genders over 18 years of age who had underwent endonasal sinus surgery because of CRS 2 weeks earlier were eligible for screening for this clinical trial. Primary surgical cases and revision cases were included, and all patients had undergone unsuccessful conservative treatment prior the operation. The diagnosis of CRS was established by the patient’s history, nasal endoscopy, and CT of the paranasal sinuses, according to the EPOS criteria. 15 All patients had signed an informed consent prior the participation of the trial. After a sample-size calculation, it was planned to include 70 patients: 35 to receive erythromycin and 35 to receive placebo. Exclusion criteria were single-sided CRS, cystic fibrosis, immunodeficiency, primary cilia dysfunction, hypersensitivity to macrolides, liver or renal deficiency, cardiac arrhythmia, QTtime elongation on electrocardiogram, pregnancy, and concomitant medication with drugs interacting with macrolides.
Study Design This monocenter prospective randomized, double-blind, placebo-controlled trial was conducted by the department of Otorhinolaryngology, Head and Neck Surgery of the Johannes
Laryngoscope 00: Month 2014
2
Examination was performed 2 weeks after surgery (visit 1), 12 weeks (visit 2), and 24 weeks (visit 3) thereafter. Neither the investigators nor the patients were aware of the treatment group into which the patients were randomized.
Follow-up Assessments Analysis of ECP and MPO in Nasal Secretions. Nasal secretions were collected by placing a sponge in both middle nasal meati. After 20 minutes, the sponges were removed and centrifuged for 5 minutes. The secretions were stored frozen and sequentially defrosted for further analysis. The ECP and MPO concentrations were measured with the ImmunoCAP method (Phadia, Freiburg, Germany). Sino-Nasal Outcome Test. For evaluation of the diseasespecific quality of life, the SNOT-20 score was used.16 This score addresses 20 items, and patients are asked to rate each item between 0 (no problem) and 5 (as bad as it can be) points. The entire score can reach values between 0 to 100 points. The German adaptation version was used in this trial.17 Olfaction. To evaluate the sense of smell, the Sniffin0 Sticks identification test was used, a standardized, 16-item, forced-choice odor identification test.18 The correct answers were noted; results between 0 and 16 were possible. Saccharin Transit Time. To evaluate the nasal ciliary clearance, the saccharin transit time (STT) was used. A small
Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
TABLE I. Demographic Data of Erythromycin and Placebo Groups (ITT population) Before Administration of Study Medication 6 Standard Error of Mean. Erythromycin
Placebo
Number
29
29
Gender Male
17 (59%)
17 (59%)
Female Age Revision cases
P Values
1.000 12 (41%)
12 (41%)
45.7 (6 12.8) 16 (55%)
47.7 (6 12.5) 16 (55%)
0.5451 1.000
CT score (Lund and Mackay)
13.9 (6 5.4)
15.8 (6 5.1)
0.193
CRSwNP blood IgE-concentration (ml/l)
15/29 (52%) 165.0 (6 195.2)
17/29 (59%) 126.9 (6 205.8)
0.605 0.528
Baseline concentration of ECP
176.4
186.9
0.637
Baseline concentration of MPO SNOT-20 total score
0.8 21.5 (6 13.7)
0.4 19.7 (6 9.3)
0.299 0.555
Subjective nasal health (VAS) STT (min) Sniffin’ Sticks total score Endoscopy score Positive testing for allergies
4.0 (6 2.4)
3.7 (6 2.5)
0.624
25.3 (6 14.9) 10.5 (6 3.0)
27.3 (6 19.3) 10.3 (6 3.7)
0.924 0.830
2.6 (6 1.4)
2.5 (6 1.3)
0.702
19 (66%)
17 (59%)
0.596
CRSwNP 5 cases with nasal polyps; CT 5computed tomography; ECP 5 eosinophilic cationic protein; IgE 5 immunoglobulinE; ITT 5 intention to treat; MPO 5 myeloperoxidase; VAS 5 visual analogue scale; STT 5saccharin transit time.
particle of saccharine was placed unilaterally on the anterior end of the inferior turbinate. The time between the placement and the first reported perception of sweetness from the patient was recorded. Subjective Severity of the Nasal Disease. According to the international guidelines, the patients were asked “How troublesome are your symptoms of rhinosinusitis?” A visual analogue scale (VAS) 10-centimeters long was used in which the patients had to mark the extent of their overall nasal condition with a score of 0 representing “not troublesome” and 10 “worst thinkable troublesome.” The score was adapted to the 2012 EPOS guidelines; and the complaints were then divided into mild, moderate, or severe. Endoscopic Assessment. Rigid nasal endoscopic assessment was classified based on three different scores: mucosal swelling (0 5 no swelling, 1 5 mild swelling, 2 5 severe swelling), mucosa colour (0 5 pale, 1 5 red), secretion (0 5 normal, 1 5 watery, 2 5 mucoid, 3 5 purulent), and the appearance of polyps (0 5 absent, 1 5 present) for each side separately. The average of both sides was documented, and the maximal possible score was 7. Ethical Considerations. The study was registered with the European Clinical Trials Database (EudraCT number 2008– 005828-84), and all participants signed an informed consent. Full ethical approval was granted from the local ethics committee. Statistical Analysis. For statistical analysis, SAS 9.2 (SAS Institute, Cary, NC) was used. The sample-size calculation was based on changes of the IL-8 concentration in the Wallwork publication.13 In this study, the standardized effect size was 0.79 (mean difference of 148, pooled standard deviation: 187). Therefore, a conservative approach, with an assumed effect size of 0.75 of significance of 5% and a two-sided t test in 2 3 29 patients, revealed a power of 80%. As the assumed dropout rate was small; a randomization of 2 3 35 patients was planned. For the primary outcomes, the level of significance that was chosen was 0.05. For statistical analysis, the differences between visit
Laryngoscope 00: Month 2014
2 and 1 and visit 3 and 1, respectively, were assessed by t tests. For the secondary parameters and analysis of subgroups, Wilcoxon tests, t tests, and analysis of covariance were used.
RESULTS From September 2009 until June 2012, 67 patients were screened, and 58 patients were included in the clinical study and randomized either to erythromycin or placebo treatment. The initially planned number of 70 patients could not be reached in an appropriate time; therefore, a premature termination of the study was decided. Thus, at visit 1, 29 patients were allocated to the erythromycin group and 29 patients to the placebo group and served as intention-to-treat population (ITT). The two groups revealed comparable baseline and clinical characteristics (Table I). During the 12 weeks with study-specific medication, one patient in each group refused further participation at the trial and one patient in the placebo group and six in the erythromycin group were lost to follow-up. One patient in the erythromycin group was diagnosed with Churg-Strauss syndrome and stopped the medication thereafter. In the following period of 12 weeks of continuous follow-up after the study specific medication, 27 patients in the placebo group and 21 in the erythromycin group could be analyzed (Fig. 1). The per protocol (PP)-population was defined as patients who have received the study medication for at least 2 months, which was the case in 29 patients in the placebo group and 26 in the erythromycin group.
Primary Parameters The initial concentration of ECP in the erythromycin group at visit 1 was 176.4 ml/l 6 79.0 and 186.9 ml/l 6 36.0 Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
3
Fig. 1. CONSORT 2010 Flow Diagram. AE 5 adverse events; ESS 5 endonasal sinus surgery; ITT 5 intention to treat; PP 5 per protocol; SAE 5 severe adverse events. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
in the placebo group. At visit 2, the difference to visit 1 could be calculated for 14 patients in the erythromycin group and showed a mean value of 151.7 ml/l 6 202.1. In the placebo group (values for 12 patients), the difference was 119.9 ml/l 6 229.9, P 5 0.7129. At visit 3, the difference to visit 1 was calculated in 11 patients in the erythromycin group (129.7 ml/l 6 220.2) and in 12 patients in the placebo group (143.5 ml/l 6 188.7, n 5 12). During all visits, no statistical difference of the ECP concentration was seen between groups. The analysis of the PP population showed no differing results compared to the ITT analysis (Fig. 2A). MPO concentrations were initially similar in the two groups; the MPO values during the following visits could only be evaluated in less than 50%. Therefore, the difference between visit 2 and visit 1 in the erythromycin group could only be calculated in four patients (20.5 U/ ml 6 0.8) and in five in the placebo group (0.5 U/ml 6 1.3) Laryngoscope 00: Month 2014
4
(P 5 0.2047) (Fig. 2B). For detailed information on all secondary outcome parameters, please see Supp. Table SII.
Secondary Parameters Sino-Nasal Outcome Test. During the study, the values of the SNOT-20 scores showed no statistical difference between both treatment groups. The mean change of the score from visit 1 to visit 2 was 23.7 6 10.4 in the erythromycin group and 25.3 6 12.6 in the placebo group (Fig. 3A) (P 5 0.6230). Olfaction. The results of the Sniffin’ Sticks tests were comparable at visit 1 (10.5 6 3.0 points in the erythromycin group and 10.3 6 3.7 points in the placebo group). In both groups, an increase could be detected in the following visits (Fig. 3B), and no statistical differences could be observed (P 5 0.7762). Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
Fig. 2. Changes in the primary outcomes from visit 1 (start of the medication 2 weeks after sinus surgery) to visit 2 (end of 12 weeks study medication) and visit 3 (follow-up of 12 weeks after ending the medication) for A (ECP levels) and B (MPO levels for the erythromycin group [blue line] and the placebo group [green line]) (values 6 standard deviation). ECP 5 eosinophilic cationic protein; MPO 5 myeloperoxidase. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
Saccharin Transit Time. In the erythromycin group, the mean STT was 25.3 6 14.9 minutes at visit 1 and 27.3 6 19.3 minutes in the placebo group. During the follow-up, the STT decreased similarly in both groups (Fig. 3C). Statistically, no relevant differences between the treatments could be detected (P 5 0.8631). Subjective Severity of the Nasal Disease. This parameter was evaluated by a VAS (0–10). The lower the value, the better was the patient’s subjective nasal
health status. The decrease of this parameter between visit 1 and 2 was 1.7 6 2.0 in the placebo group and 0.8 6 3.5 in the erythromycin group, but the difference was not statistically significant (P 5 0.3267) (Fig. 3D). Endoscopic Assessment. Nasal endoscopy scores could reach values between 0 and 7. The initial mean values at visit 1 were comparable between the two treatment groups (2.6 6 1.4 in the erythromycin group and 2.5 6 1.3 in the placebo group). At visit 2, the mean
Fig. 3. Changes in the secondary outcomes from visit 1 (start of the medication 2 weeks after sinus surgery) to visit 2 (end of 12 weeks study medication) and visit 3 (followup of 12 weeks after ending the medication) for A (SNOT-20 score), B (STT), C (Sniffin’ Sticks score), D (severity score [visual analogue scale]), the erythromycin group (blue line), and the placebo group (green line) (values 6 standard deviation). SNOT 5 Sino-Nasal Outcome Test; STT 5saccharin transit time. [Color figure can be viewed in the online issue, which is available at www. laryngoscope.com.]
Laryngoscope 00: Month 2014
Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
5
in 5 patients). Theses AEs lead to a discontinuation of the study in four patients in the erythromycin group and one patient in the placebo group (Fig. 1). Subgroup Analysis. For a subgroup analysis, different parameters have been investigated: the patients’ blood IgE-level (< 100 5 low, > 100 5 high), the CT score after Lund-Mackay (< 12 points 5 low, > 12 points 5 high), atopic disposition in prick or radioallergosorbent test (RAST) (0 5 no atopy, 1 5 atopy), and nasal polyposis (0 5 none, 1 5 present). None of the parameters showed a statistically significant advantage between the subgroups over placebo except for the patients with CRSwNP compared to CRSsNP; in the latter subgroup, a trend toward an improvement in erythromycin-treated patients in all secondary outcome criteria was detected (Fig. 5). Nevertheless, except the endoscopy score (P 5 0.006), these results failed statistical significance (P value for SNOT20 5 0.076, P value for Sniffin’ Sticks 5 0.147, P value for VAS 5 0.218, P value for STT 5 0.794).
DISCUSSION Fig. 4. Changes in the secondary outcome endoscopy score from visit 1 (start of the medication 2 weeks after sinus surgery) to visit 2 (end of 12 weeks study medication) and visit 3 (follow-up of 12 weeks after ending the medication) for the erythromycin group (blue line) and the placebo group (green line) (values 6 standard deviation). * indicates significant changes (P < 0.05). [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
value in the erythromycin group dropped to 1.9 6 1.6, whereas it remained stable in the placebo group (2.6 6 1.5). The differences between the values of visit 2 and visit 1 varied significantly in the two treatment groups (P 5 0.0351). At visit 3, the two groups approximated at a lower level (1.9 6 1.6 for the erythromycin group and for 2.0 6 1.5 the placebo group), showing no statistical difference (Fig. 4). For detailed information on all secondary outcome parameters, please see Supp. Table SII. Side Effects. In each group, two severe adverse events (SAE) occurred during the clinical trial. In the placebo group, one event of epistaxis resulted in hospital admission for conservative treatment (packing) and resolved completely. One patient suffered from acute dyspnea, which also resulted in hospital admission but settled after medical treatment. In the erythromycin group, a patient was also admitted for conservative treatment of epistaxis, and another patient was admitted to a pulmonary unit with newly diagnosed ChurgStrauss syndrome. Episodes of an acute bacterial infection of the upper respiratory tract (acute sinusitis, bronchitis, otitis, or bronchitis) occurred in both treatment groups, with a higher incidence in the erythromycin group (18 incidences in 13 patients vs. 10 in 9 patients in the placebo group). Adverse events (AE) such as gastrointestinal disorders occurred only in the erythromycin group (6 events Laryngoscope 00: Month 2014
6
At the time that this trial was designed and started (2008 and 2009, respectively), the knowledge about macrolides and CRS was quite limited, including the recommendations in the 2007 EPOS guidelines.19 Only one prospective, randomized study of long-term erythromycin therapy (250 mg twice daily) versus surgical treatment in a mixed population (polypoid and nonpolypoid CRS) was available that showed no significant changes between the two treatment groups in different clinical parameters. Those with polyps even showed a trend to a pronounced improvement in certain subjective parameters. Only one DBPC study in a population of CRSsNP was available that showed a significant improvement in the macrolide-treated group.13 Therefore, in our study, CRS patients with and without nasal polyps were included to investigate the effects of a macrolide treatment. However, with the knowledge available today and the negative results from another DPBC study in a mixed population,14 the selection would have been different. A recent meta-analysis20 also identified the two above-mentioned controlled studies,13,14 as well as another study for which clarithromycin was administered for 3 weeks only and compared to amoxicillin/clavulanic acid21 as prospective trials. The authors identified that there is only a statistically significant difference favoring the macrolide therapy at one (of three different) time point (12 weeks after discontinuation of a 12-week medication) in the SNOT-scores (standard mean difference of 20.43 [20.82, 20.05]). All other investigated parameters failed to show significance. However, the clinical relevance of this difference in the SNOT score is doubtful. The present study is the third randomized, doubleblind, placebo-controlled trial on macrolides in chronic rhinosinusitis and the first after a surgical intervention. One multicenter study with 60 patients and a medication of azithromycin revealed no advantage over Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
Fig. 5. Results of the percentual changes in different parameters between visit 1 and visit 2 in the subgroups of CRSwNP and CRSsNP in the ERT and PBO groups. * indicates significant changes (P < 0.05). CRSsNP 5 cases without nasal polyposis; CRSwNP 5 cases with nasal polyps; ERT 5erythromycin; PBO 5 placebo; SNOT 5 Sino-Nasal Outcome Test; VAS 5 visual analogue scale; ENDO 5 endoscopy score. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
placebo.14 Only the parameter facial pain improved significantly in the azithromycin group. The DBPC monocenter Wallwork et al. trial, with 64 patients using roxithromycin over a period of 12 weeks, showed a higher reduction of the mean symptom response score compared to the placebo group, as well as a more pronounced and statistically significant change in the SNOT-20 score from baseline in the roxithromycin group at the end of treatment (0.46 points on a scale of 0 to 5).13 In this study, especially the subgroup of patients with low serum IgE showed significant improvements in the saccharine transit time and nasal endoscopic score and SNOT-20 score after 12 weeks, which could not be proven in our study. Similar to the results of the Wallwork et al. study13 with only CRSsNP patient included, in our study, the subgroup of CRSsNP patients also showed a tendency to a greater improvement after the macrolide therapy in most of the evaluated parameters (Fig. 5). It must be admitted that some methodological problems have occurred in our study: 1. The initially planned number of 70 participants could not be achieved, and the study was prematurely closed because the recruitment of patients could not be completed in a reasonable period of time. The study medication would have to be revolved after 3 years, and the costs for this action were too extensive in a nonfunded investigator-initiated trial. Therefore, the study unfortunately lost power. 2. The primary outcome criteria with concentrations of ECP and MPO in nasal secretions were initially chosen with the intention to receive objective results. In previous studies with long-term, low-dose clarithromycin in CRS patients, ECP as well as MPO levels in nasal secretions decreased
Laryngoscope 00: Month 2014
under the treatment,10 similar to a study with doxycycline treatment.22 These observations could not been reproduced in our trial; a reason might be the surgical trauma influencing the inflammatory markers substantially in our study. After surgical interventions, an inflammatory mucosal reaction was described, leading to an increase of inflammatory markers in nasal secretions. In patients suffering from cystic fibrosis who had surgery for CRS, ECP levels in nasal secretions stayed high or even increased after 1 month and declined only 4 months after surgery.23 The inflammatory reaction seems to outweigh possible beneficial effects of the macrolide treatment. Additionally, the volume of the nasal secretions gained by the sponge method that was used was too low, and results for MPO could only be gathered in a small amount of patients. Therefore, the initially planned determination of the concentrations of other specific markers (IL-8 and IL-5) also had to be abandoned completely. 3. The surgical intervention itself might have influenced the course of action so strongly that a possible additional effect of a macrolide therapy was covered. One study compared a medical treatment for CRS (with and without nasal polyps) with macrolides (loading dose of erythromycin 500 mg twice daily for 2 weeks, followed by 250 mg twice daily for 10 weeks) to a surgical intervention.24 Here, a significant increase in subjective and objective parameters in both groups could be found up to 1 year of follow-up without significant differences between the groups despite nasal volume, which was in favor for the surgical group. In a recently published review,25 additional controlled trials also in a postoperative setting are proposed.
Nevertheless, in our study with low-dose erythromycin 250 mg per day over a period of 3 months, the nasal endoscopy score showed a significant improvement in erythromycin-treated patients over placebo during the Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
7
study period. The amount of the difference of 1 point (out of 7 possible) also seems to be clinically relevant. In the follow-up period of an additional 3 months, this endoscopy score also decreased to lower values in the placebo group (Fig. 4). The other evaluated secondary parameters (including SNOT-20 score, self-rating of nasal health on a VAS, olfaction, and saccharin transit time) decreased in both groups over time and showed no statistically significant differences between the two treatment groups. Additionally, no positive effect on the primary criteria (ECP and MPO in nasal secretions) was found. In our study, no severe side effects were seen, and the study medication did not cause higher rates of severe adverse events compared to placebo. These results are in accordance with previous reports.13,26,27 But the ratio of side effects, especially of the gastrointestinal tract, was higher in the erythromycin group, which lead to a higher number of dropouts. Therefore, a higher dosage as used in other studies might cause even more side effects. It might well be possible that a macrolide medication over 12 weeks, as used in our study, is not long enough. As in another nonplacebo-controlled postsurgical study, the visual analogue scale scores for rhinorrhea and postnasal drip at 12 months after surgery were lower in a group that was treated with clarithromycin (200 mg/day) for 6 months than in the group with 3 months of treatment.28 One general problem in these types of studies is the limited number of patients who can be included. Both previously published DBPC studies showed similar patient numbers to the present study, one even in a multicenter design. Therefore, the results must always be discussed with caution because more included patients would mean more statistical power. On the other hand, investigator-initiated trials often are confronted with critical funding and duration is limited.
CONCLUSION In conclusion, the nasal markers for inflammation (ECP and MPO) could not be influenced by a long-term, low-dose erythromycin therapy in patients after sinus surgery. Despite a positive effect on the nasal endoscopy score in the treatment group, no other clinical advantage of the study medication over placebo could be proven. Therefore a general recommendation for the use of longterm, low-dose erythromycin therapy after surgery for CRS cannot be advocated. A presumably relevant improvement in the subgroup of patients without nasal polyps (CRSsNP) and specific endotypes (neutrophilic inflammation) should be further investigated.
BIBLIOGRAPHY 1. Fokkens WJ, Lund VJ, Mullol J, et al. EPOS 2012: European position paper on rhinosinusitis and nasal polyps 2012. A summary for otorhinolaryngologists. Rhinology 2012;50:1–12.
Laryngoscope 00: Month 2014
8
2. Bachert C, Zhang N, van Zele T, Gevaert P, Patou J, van Cauwenberge P. Staphylococcus aureus enterotoxins as immune stimulants in chronic rhinosinusitis. Clin Allergy Immunol 2007;20:163–175. 3. Gosepath J, Pogodsky T, Mann WJ. Characteristics of recurrent chronic rhinosinusitis after previous surgical therapy. Acta Otolaryngol 2008; 128:778–784. 4. Ponikau JU, Sherris DA, Kita H. The role of ubiquitous airborne fungi in chronic rhinosinusitis. Clin Allergy Immunol 2007;20:177–184. 5. Van Zele T, Claeys S, Gevaert P, et al. Differentiation of chronic sinus diseases by measurement of inflammatory mediators. Allergy 2006;61: 1280–1289. 6. Zhang N, Van Zele T, Perez-Novo C, et al. Different types of T-effector cells orchestrate mucosal inflammation in chronic sinus disease. J Allergy Clin Immunol 2008;122:961–968. 7. Smith TL, Litvack JR, Hwang PH, et al. Determinants of outcomes of sinus surgery: a multi-institutional prospective cohort study. Otolaryngol Head Neck Surg 2010;142:55–63. 8. Mendelsohn D, Jeremic G, Wright ED, Rotenberg BW. Revision rates after endoscopic sinus surgery: a recurrence analysis. Ann Otol Rhinol Laryngol 2011;120:162–166. 9. Wallwork B, Coman W. Chronic rhinosinusitis and eosinophils: do macrolides have an effect? Curr Opin Otolaryngol Head Neck Surg 2004;12: 14–17. 10. Cervin A, Wallwork B, Mackay-Sim A, Coman WB, Greiff L. Effects of long-term clarithromycin treatment on lavage-fluid markers of inflammation in chronic rhinosinusitis. Clin Physiol Funct Imaging 2009;29: 136–142. 11. Wallwork B, Coman W, Feron F, Mackay-Sim A, Cervin A. Clarithromycin and prednisolone inhibit cytokine production in chronic rhinosinusitis. Laryngoscope 2002;112:1827–1830. 12. Elmorsy S, El-Naggar MM, Abdel aal SM, Abou-elela MA. Sinus aspirates in chronic rhinosinusitis: fungal colonization of paranasal sinuses, evaluation of ICAM-1 and IL-8 and studying of immunological effect of longterm macrolide therapy. Rhinology 2010;48:312–317. 13. Wallwork B, Coman W, Mackay-Sim A, Greiff L, Cervin A. A double-blind, randomized, placebo-controlled trial of macrolide in the treatment of chronic rhinosinusitis. Laryngoscope 2006;116:189–193. 14. Videler WJ, Badia L, Harvey RJ, et al. Lack of efficacy of long-term, lowdose azithromycin in chronic rhinosinusitis: a randomized controlled trial. Allergy 2011;66:1457–1468. 15. Fokkens W, Lund V, Mullol J. European position paper on rhinosinusitis and nasal polyps 2007. Rhinol Suppl 2007:1–136. 16. Piccirillo JF, Merritt MG Jr, Richards ML. Psychometric and clinimetric validity of the 20-item sino-nasal outcome test (SNOT-20). Otolaryngol Head Neck Surg 2002;126:41–47. 17. Baumann I. [Validated instruments to measure quality of life in patients with chronic rhinosinusitis]. HNO 2009;57:873–881. 18. Wolfensberger M, Schnieper I, Welge-Lussen A. Sniffin’Sticks: a new olfactory test battery. Acta Otolaryngol 2000;120:303–306. 19. Fokkens W, Lund V, Mullol J. EP3OS 2007: European position paper on rhinosinusitis and nasal polyps 2007. A summary for otorhinolaryngologists. Rhinology 2007;45:97–101. 20. Pynnonen MA, Venkatraman G, Davis GE. Macrolide therapy for chronic rhinosinusitis: a meta-analysis. Otolaryngol Head Neck Surg 2013;148: 366–373. 21. Amini M, Yarmohammadi M, Izadi P. A comparing study of clarithromycin XL with co-amoxiclav for treatment of chronic sinusitis; a clinical trial. Iranian Journal of Clinical Infectious Diseases 2009;4:197–201. 22. Van Zele T, Gevaert P, Holtappels G, et al. Oral steroids and doxycycline: two different approaches to treat nasal polyps. J Allergy Clin Immunol 2010;125:1069–1076 e1064. 23. Weber A, Kiefer J, Peters S, Schneider M, Bargon J, May A. Eosinophilic cationic protein as a marker of nasal inflammation in patients with cystic fibrosis. Laryngoscope 1999;109:1696–1702. 24. Ragab SM, Lund VJ, Scadding G. Evaluation of the medical and surgical treatment of chronic rhinosinusitis: a prospective, randomized, controlled trial. Laryngoscope 2004;114:923–930. 25. Cervin A, Wallwork B. Efficacy and safety of long-term antibiotics (macrolides) for the treatment of chronic rhinosinusitis. Curr Allergy Asthma Rep 2014;14:416. 26. Cervin A, Wallwork B. Macrolide therapy of chronic rhinosinusitis. Rhinology 2007;45:259–267. 27. Videler WJ, van Hee K, Reinartz SM, Georgalas C, van der Meulen FW, Fokkens WJ. Long-term low-dose antibiotics in recalcitrant chronic rhinosinusitis: a retrospective analysis. Rhinology 2012;50:45–55. 28. Nakamura Y, Suzuki M, Yokota M, et al. Optimal duration of macrolide treatment for chronic sinusitis after endoscopic sinus surgery. Auris Nasus Larynx 2013;40:366–372.
Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
Rhinological and Otological Society, Inc.
Controlled Trial for Long-Term Low-Dose Erythromycin After Sinus Surgery for Chronic Rhinosinusitis Boris R. Haxel, MD, PhD; Meike Clemens; Niki Karaiskaki, MD; Uta Dippold, MD; Lisanne Kettern, MD; Wolf J. Mann, MD, PhD, FACS Objectives/Hypothesis: The efficacy of macrolides in chronic rhinosinusitis (CRS) is still under controversy. To date, only two double-blind, placebo-controlled studies have been published with differing results. None of these studies investigated the possible benefit of macrolides in the postoperative period. We conducted an investigator-initiated clinical trial using 250-mg erythromycin once a day over a period of 3 months, beginning the administration of either erythromycin or placebo 2 weeks after a surgical intervention for CRS. Study Design: Randomized double-blind, placebo-controlled trial. Methods: The concentrations of eosinophilic cationic protein (ECP) and myeloperoxidase in nasal secretion were chosen as primary outcome measures. Additionally, as a secondary outcome measure, changes in the Sino-Nasal Outcome Test-20 score, olfaction, saccharin transit time, nasal endoscopy score, and self-rating of nasal health using a visual analogue scale were evaluated. Results: Sixty-seven patients after surgery for CRS with or without nasal polyps were screened, and 58 patients were randomized to the study groups. For the primary outcomes, the concentrations of ECP changed from 176.4 ml/l 6 79.0 to 226.1 ml/l 6 200.6 in the erythromycin group and from 186.9 ml/l 6 36.0 to 192.9 ml/l 6 189.2 in the placebo group; no statistical differences were found. Of the secondary outcomes, only the nasal endoscopy score showed a statistically significant improvement in the erythromycin group (from 2.6 6 1.4 to 1.9 6 1.5 points) compared to the placebo group (from 2.5 6 1.3 to 2.6 6 1.5 points). The subgroup of patients without nasal polyps in the erythromycin group showed a tendency to improvement in some secondary outcome criteria. Conclusions: A general recommendation for long-term, low-dose erythromycin treatment after surgery for CRS cannot be given. In patients with CRS without nasal polyps, a tendency to improved parameters was detected. Key Words: Macrolides, sinus surgery, CRS, evidence-based medicine. Level of Evidence: Ib. Laryngoscope, 00:000–000, 2014
INTRODUCTION Rhinosinusitis is regarded as a multifactorial disease that causes different symptoms such as nasal obstruction, headache, nasal discharge, and olfactory dysfunction. Chronic rhinosinusitis (CRS) is defined as a state of inflammatory disease that occurs for longer than 12 weeks and is diagnosed by nasal endoscopy and computed tomography (CT) scan of the sinuses. CRS can
Additional Supporting Information may be found in the online version of this article. From the Department of Otorhinolaryngology, Head and Neck Surgery, AMEOS Klinikum Haldensleben (B.R.H.), Haldensleben; and the Department of Otorhinolaryngology, Head and Neck Surgery, University Medical Center of the Johannes Gutenberg University (B.R.H., M.C., N.K., U.D., L.K., W.J.M.), Mainz, Germany. Editor’s Note: This Manuscript was accepted for publication October 29, 2014. Presented at the 84th Annual Meeting of the German Society of Oto-Rhino-Laryngology, Head and Neck Surgery, Nuremberg, Germany, 2013. The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Boris R. Haxel, MD, PhD, Department of Otorhinolaryngology, Head and Neck Surgery, AMEOS Klinikum Haldensleben, Kiefholzstr 27, D-39340 Haldensleben, Germany. E-mail: [email protected] DOI: 10.1002/lary.25052
Laryngoscope 00: Month 2014
present as cases with nasal polyps (CRSwNP) and without nasal polyps (CRSsNP), which seem to have a different underlying pathophysiology.1 Possible causes for the inflammation in the nasal mucosa include staphylococcal superantigen, aspirin-exacerbated respiratory disease, biofilm, or fungi, just to mention the most common.2–4 Present and future research efforts aim to find differences in individual phenotypes and endotypes such as cytokine levels in nasal secretion to be able to distinguish between different specifications of the underlying inflammatory disease.5,6 Therapeutic options are mainly topical nasal steroids, antibiotics in acute exacerbations, and short-term oral steroids. In patients who do not improve after a conservative treatment, a surgical approach (functional endonasal sinus surgery) is recommended.1 Surgery is effective to control local inflammation and improves the quality of life7 but is usually followed by long-term topical steroid treatment. Especially in CRSwNP patients, the risk of recurrent disease is higher than in CRSsNP patients, reaching rates of about 20%.8 In some studies, long-term, low-dose macrolides are proposed for medical treatment of CRS.9,10 Here, macrolides are thought to have antiinflammatory or immunomodulatory capacities besides their antimicrobial effects. Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
1
This is based on the observation of blockage of the production of interleukin-8 (IL-8) and tumor necrosis factor, the effects on neutrophil migration and adhesion, and modulation of synthesis and secretion of mucus.11 In clinical trials of CRS, macrolides showed improvements in symptoms and endoscopic findings, decrease in polyp size, decrease of radiologically conspicuous mucosal changes in the sinuses on CT scan, and reduction in neutrophils and interleukin (IL)28 levels in nasal discharge—as well as an improvement in saccharine transit time.10,12 Although these results showed promising effects of macrolides in CRS, most of these studies were conducted as open and noncontrolled clinical trials. To date, only two double-blind, placebo-controlled (DBPC) studies have been published with differing results.13,14 In the Wallwork study, CRS patients without topical or systemic treatment 4 weeks prior to the start of the macrolide therapy were included; whereas in the Videler study, patients who suffered from recalcitrant CRS despite previous surgery and intensive medical treatment were evaluated.13,14 None of these studies investigated the possible benefit of macrolides in the immediate postoperative period. In the 2012 European Position Paper on Rhinosinusitis (EPOS) guidelines, macrolides for treatment of CRS is recommended for patients with CRSsNP (level of evidence Ib, strength of recommendation C).1 We conducted a randomized, double-blind, placebocontrolled clinical trial with low-dose erythromycin over a period of 3 months directly after a surgical intervention for CRS. The changes in concentrations of eosinophilic cationic protein (ECP) and myeloperoxidase (MPO) in nasal secretion were chosen as primary outcome measures. Additionally, we followed clinically and functional parameters and possible side effects of the study medication.
Gutenberg University, Mainz, Germany, as an investigatorinitiated study.
Study Medication As study medication, 250-mg erythromycin once a day or placebo was administered over a period of 3 months, beginning 2 weeks after the surgical intervention for CRS. A randomization list was generated by a computer program in the Interdisciplinary Center for Clinical Trials and was confidentially stored in this institution. A copy of this list was only forwarded to the pharmacy of the university hospital. The pharmacy provided the packing and blinding of the study medication. The erythromycin 250-mg tablets were packed into capsules. Identical empty capsules served as placebo.
Concomitant Medication All patients used nasal irrigation with saline at least twice a day and were treated with a topical nasal steroid (fluticasone furoate) once a day (27.5 mg), beginning after the surgical intervention and continuing for the entire study period.
Primary and Secondary Outcome Parameters The concentrations of ECP and MPO in the nasal secretions were chosen as primary outcomes. Additionally, changes in the nasal endoscopy score, saccharin transit time, olfaction, Sino-Nasal Outcome Test (SNOT-20) score, and a subjective nasal health status determined by a visual analogue scale were evaluated. Additionally, the preoperative CT scans of the paranasal sinuses were analyzed and the Lund-Mackay score was noted, as well as the status of atopy by using a skin prick test and in vitro methods (specific immunoglobulin E [IgE] for the most common environmental allergens in the ImmunoCAP testing). Furthermore, the blood IgE-levels of the patients were determined, and it was noted whether the patient suffered from CRSsNP or CRSwNP.
Follow-up MATERIALS AND METHODS Patients In a prospective study, patients of both genders over 18 years of age who had underwent endonasal sinus surgery because of CRS 2 weeks earlier were eligible for screening for this clinical trial. Primary surgical cases and revision cases were included, and all patients had undergone unsuccessful conservative treatment prior the operation. The diagnosis of CRS was established by the patient’s history, nasal endoscopy, and CT of the paranasal sinuses, according to the EPOS criteria. 15 All patients had signed an informed consent prior the participation of the trial. After a sample-size calculation, it was planned to include 70 patients: 35 to receive erythromycin and 35 to receive placebo. Exclusion criteria were single-sided CRS, cystic fibrosis, immunodeficiency, primary cilia dysfunction, hypersensitivity to macrolides, liver or renal deficiency, cardiac arrhythmia, QTtime elongation on electrocardiogram, pregnancy, and concomitant medication with drugs interacting with macrolides.
Study Design This monocenter prospective randomized, double-blind, placebo-controlled trial was conducted by the department of Otorhinolaryngology, Head and Neck Surgery of the Johannes
Laryngoscope 00: Month 2014
2
Examination was performed 2 weeks after surgery (visit 1), 12 weeks (visit 2), and 24 weeks (visit 3) thereafter. Neither the investigators nor the patients were aware of the treatment group into which the patients were randomized.
Follow-up Assessments Analysis of ECP and MPO in Nasal Secretions. Nasal secretions were collected by placing a sponge in both middle nasal meati. After 20 minutes, the sponges were removed and centrifuged for 5 minutes. The secretions were stored frozen and sequentially defrosted for further analysis. The ECP and MPO concentrations were measured with the ImmunoCAP method (Phadia, Freiburg, Germany). Sino-Nasal Outcome Test. For evaluation of the diseasespecific quality of life, the SNOT-20 score was used.16 This score addresses 20 items, and patients are asked to rate each item between 0 (no problem) and 5 (as bad as it can be) points. The entire score can reach values between 0 to 100 points. The German adaptation version was used in this trial.17 Olfaction. To evaluate the sense of smell, the Sniffin0 Sticks identification test was used, a standardized, 16-item, forced-choice odor identification test.18 The correct answers were noted; results between 0 and 16 were possible. Saccharin Transit Time. To evaluate the nasal ciliary clearance, the saccharin transit time (STT) was used. A small
Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
TABLE I. Demographic Data of Erythromycin and Placebo Groups (ITT population) Before Administration of Study Medication 6 Standard Error of Mean. Erythromycin
Placebo
Number
29
29
Gender Male
17 (59%)
17 (59%)
Female Age Revision cases
P Values
1.000 12 (41%)
12 (41%)
45.7 (6 12.8) 16 (55%)
47.7 (6 12.5) 16 (55%)
0.5451 1.000
CT score (Lund and Mackay)
13.9 (6 5.4)
15.8 (6 5.1)
0.193
CRSwNP blood IgE-concentration (ml/l)
15/29 (52%) 165.0 (6 195.2)
17/29 (59%) 126.9 (6 205.8)
0.605 0.528
Baseline concentration of ECP
176.4
186.9
0.637
Baseline concentration of MPO SNOT-20 total score
0.8 21.5 (6 13.7)
0.4 19.7 (6 9.3)
0.299 0.555
Subjective nasal health (VAS) STT (min) Sniffin’ Sticks total score Endoscopy score Positive testing for allergies
4.0 (6 2.4)
3.7 (6 2.5)
0.624
25.3 (6 14.9) 10.5 (6 3.0)
27.3 (6 19.3) 10.3 (6 3.7)
0.924 0.830
2.6 (6 1.4)
2.5 (6 1.3)
0.702
19 (66%)
17 (59%)
0.596
CRSwNP 5 cases with nasal polyps; CT 5computed tomography; ECP 5 eosinophilic cationic protein; IgE 5 immunoglobulinE; ITT 5 intention to treat; MPO 5 myeloperoxidase; VAS 5 visual analogue scale; STT 5saccharin transit time.
particle of saccharine was placed unilaterally on the anterior end of the inferior turbinate. The time between the placement and the first reported perception of sweetness from the patient was recorded. Subjective Severity of the Nasal Disease. According to the international guidelines, the patients were asked “How troublesome are your symptoms of rhinosinusitis?” A visual analogue scale (VAS) 10-centimeters long was used in which the patients had to mark the extent of their overall nasal condition with a score of 0 representing “not troublesome” and 10 “worst thinkable troublesome.” The score was adapted to the 2012 EPOS guidelines; and the complaints were then divided into mild, moderate, or severe. Endoscopic Assessment. Rigid nasal endoscopic assessment was classified based on three different scores: mucosal swelling (0 5 no swelling, 1 5 mild swelling, 2 5 severe swelling), mucosa colour (0 5 pale, 1 5 red), secretion (0 5 normal, 1 5 watery, 2 5 mucoid, 3 5 purulent), and the appearance of polyps (0 5 absent, 1 5 present) for each side separately. The average of both sides was documented, and the maximal possible score was 7. Ethical Considerations. The study was registered with the European Clinical Trials Database (EudraCT number 2008– 005828-84), and all participants signed an informed consent. Full ethical approval was granted from the local ethics committee. Statistical Analysis. For statistical analysis, SAS 9.2 (SAS Institute, Cary, NC) was used. The sample-size calculation was based on changes of the IL-8 concentration in the Wallwork publication.13 In this study, the standardized effect size was 0.79 (mean difference of 148, pooled standard deviation: 187). Therefore, a conservative approach, with an assumed effect size of 0.75 of significance of 5% and a two-sided t test in 2 3 29 patients, revealed a power of 80%. As the assumed dropout rate was small; a randomization of 2 3 35 patients was planned. For the primary outcomes, the level of significance that was chosen was 0.05. For statistical analysis, the differences between visit
Laryngoscope 00: Month 2014
2 and 1 and visit 3 and 1, respectively, were assessed by t tests. For the secondary parameters and analysis of subgroups, Wilcoxon tests, t tests, and analysis of covariance were used.
RESULTS From September 2009 until June 2012, 67 patients were screened, and 58 patients were included in the clinical study and randomized either to erythromycin or placebo treatment. The initially planned number of 70 patients could not be reached in an appropriate time; therefore, a premature termination of the study was decided. Thus, at visit 1, 29 patients were allocated to the erythromycin group and 29 patients to the placebo group and served as intention-to-treat population (ITT). The two groups revealed comparable baseline and clinical characteristics (Table I). During the 12 weeks with study-specific medication, one patient in each group refused further participation at the trial and one patient in the placebo group and six in the erythromycin group were lost to follow-up. One patient in the erythromycin group was diagnosed with Churg-Strauss syndrome and stopped the medication thereafter. In the following period of 12 weeks of continuous follow-up after the study specific medication, 27 patients in the placebo group and 21 in the erythromycin group could be analyzed (Fig. 1). The per protocol (PP)-population was defined as patients who have received the study medication for at least 2 months, which was the case in 29 patients in the placebo group and 26 in the erythromycin group.
Primary Parameters The initial concentration of ECP in the erythromycin group at visit 1 was 176.4 ml/l 6 79.0 and 186.9 ml/l 6 36.0 Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
3
Fig. 1. CONSORT 2010 Flow Diagram. AE 5 adverse events; ESS 5 endonasal sinus surgery; ITT 5 intention to treat; PP 5 per protocol; SAE 5 severe adverse events. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
in the placebo group. At visit 2, the difference to visit 1 could be calculated for 14 patients in the erythromycin group and showed a mean value of 151.7 ml/l 6 202.1. In the placebo group (values for 12 patients), the difference was 119.9 ml/l 6 229.9, P 5 0.7129. At visit 3, the difference to visit 1 was calculated in 11 patients in the erythromycin group (129.7 ml/l 6 220.2) and in 12 patients in the placebo group (143.5 ml/l 6 188.7, n 5 12). During all visits, no statistical difference of the ECP concentration was seen between groups. The analysis of the PP population showed no differing results compared to the ITT analysis (Fig. 2A). MPO concentrations were initially similar in the two groups; the MPO values during the following visits could only be evaluated in less than 50%. Therefore, the difference between visit 2 and visit 1 in the erythromycin group could only be calculated in four patients (20.5 U/ ml 6 0.8) and in five in the placebo group (0.5 U/ml 6 1.3) Laryngoscope 00: Month 2014
4
(P 5 0.2047) (Fig. 2B). For detailed information on all secondary outcome parameters, please see Supp. Table SII.
Secondary Parameters Sino-Nasal Outcome Test. During the study, the values of the SNOT-20 scores showed no statistical difference between both treatment groups. The mean change of the score from visit 1 to visit 2 was 23.7 6 10.4 in the erythromycin group and 25.3 6 12.6 in the placebo group (Fig. 3A) (P 5 0.6230). Olfaction. The results of the Sniffin’ Sticks tests were comparable at visit 1 (10.5 6 3.0 points in the erythromycin group and 10.3 6 3.7 points in the placebo group). In both groups, an increase could be detected in the following visits (Fig. 3B), and no statistical differences could be observed (P 5 0.7762). Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
Fig. 2. Changes in the primary outcomes from visit 1 (start of the medication 2 weeks after sinus surgery) to visit 2 (end of 12 weeks study medication) and visit 3 (follow-up of 12 weeks after ending the medication) for A (ECP levels) and B (MPO levels for the erythromycin group [blue line] and the placebo group [green line]) (values 6 standard deviation). ECP 5 eosinophilic cationic protein; MPO 5 myeloperoxidase. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
Saccharin Transit Time. In the erythromycin group, the mean STT was 25.3 6 14.9 minutes at visit 1 and 27.3 6 19.3 minutes in the placebo group. During the follow-up, the STT decreased similarly in both groups (Fig. 3C). Statistically, no relevant differences between the treatments could be detected (P 5 0.8631). Subjective Severity of the Nasal Disease. This parameter was evaluated by a VAS (0–10). The lower the value, the better was the patient’s subjective nasal
health status. The decrease of this parameter between visit 1 and 2 was 1.7 6 2.0 in the placebo group and 0.8 6 3.5 in the erythromycin group, but the difference was not statistically significant (P 5 0.3267) (Fig. 3D). Endoscopic Assessment. Nasal endoscopy scores could reach values between 0 and 7. The initial mean values at visit 1 were comparable between the two treatment groups (2.6 6 1.4 in the erythromycin group and 2.5 6 1.3 in the placebo group). At visit 2, the mean
Fig. 3. Changes in the secondary outcomes from visit 1 (start of the medication 2 weeks after sinus surgery) to visit 2 (end of 12 weeks study medication) and visit 3 (followup of 12 weeks after ending the medication) for A (SNOT-20 score), B (STT), C (Sniffin’ Sticks score), D (severity score [visual analogue scale]), the erythromycin group (blue line), and the placebo group (green line) (values 6 standard deviation). SNOT 5 Sino-Nasal Outcome Test; STT 5saccharin transit time. [Color figure can be viewed in the online issue, which is available at www. laryngoscope.com.]
Laryngoscope 00: Month 2014
Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
5
in 5 patients). Theses AEs lead to a discontinuation of the study in four patients in the erythromycin group and one patient in the placebo group (Fig. 1). Subgroup Analysis. For a subgroup analysis, different parameters have been investigated: the patients’ blood IgE-level (< 100 5 low, > 100 5 high), the CT score after Lund-Mackay (< 12 points 5 low, > 12 points 5 high), atopic disposition in prick or radioallergosorbent test (RAST) (0 5 no atopy, 1 5 atopy), and nasal polyposis (0 5 none, 1 5 present). None of the parameters showed a statistically significant advantage between the subgroups over placebo except for the patients with CRSwNP compared to CRSsNP; in the latter subgroup, a trend toward an improvement in erythromycin-treated patients in all secondary outcome criteria was detected (Fig. 5). Nevertheless, except the endoscopy score (P 5 0.006), these results failed statistical significance (P value for SNOT20 5 0.076, P value for Sniffin’ Sticks 5 0.147, P value for VAS 5 0.218, P value for STT 5 0.794).
DISCUSSION Fig. 4. Changes in the secondary outcome endoscopy score from visit 1 (start of the medication 2 weeks after sinus surgery) to visit 2 (end of 12 weeks study medication) and visit 3 (follow-up of 12 weeks after ending the medication) for the erythromycin group (blue line) and the placebo group (green line) (values 6 standard deviation). * indicates significant changes (P < 0.05). [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
value in the erythromycin group dropped to 1.9 6 1.6, whereas it remained stable in the placebo group (2.6 6 1.5). The differences between the values of visit 2 and visit 1 varied significantly in the two treatment groups (P 5 0.0351). At visit 3, the two groups approximated at a lower level (1.9 6 1.6 for the erythromycin group and for 2.0 6 1.5 the placebo group), showing no statistical difference (Fig. 4). For detailed information on all secondary outcome parameters, please see Supp. Table SII. Side Effects. In each group, two severe adverse events (SAE) occurred during the clinical trial. In the placebo group, one event of epistaxis resulted in hospital admission for conservative treatment (packing) and resolved completely. One patient suffered from acute dyspnea, which also resulted in hospital admission but settled after medical treatment. In the erythromycin group, a patient was also admitted for conservative treatment of epistaxis, and another patient was admitted to a pulmonary unit with newly diagnosed ChurgStrauss syndrome. Episodes of an acute bacterial infection of the upper respiratory tract (acute sinusitis, bronchitis, otitis, or bronchitis) occurred in both treatment groups, with a higher incidence in the erythromycin group (18 incidences in 13 patients vs. 10 in 9 patients in the placebo group). Adverse events (AE) such as gastrointestinal disorders occurred only in the erythromycin group (6 events Laryngoscope 00: Month 2014
6
At the time that this trial was designed and started (2008 and 2009, respectively), the knowledge about macrolides and CRS was quite limited, including the recommendations in the 2007 EPOS guidelines.19 Only one prospective, randomized study of long-term erythromycin therapy (250 mg twice daily) versus surgical treatment in a mixed population (polypoid and nonpolypoid CRS) was available that showed no significant changes between the two treatment groups in different clinical parameters. Those with polyps even showed a trend to a pronounced improvement in certain subjective parameters. Only one DBPC study in a population of CRSsNP was available that showed a significant improvement in the macrolide-treated group.13 Therefore, in our study, CRS patients with and without nasal polyps were included to investigate the effects of a macrolide treatment. However, with the knowledge available today and the negative results from another DPBC study in a mixed population,14 the selection would have been different. A recent meta-analysis20 also identified the two above-mentioned controlled studies,13,14 as well as another study for which clarithromycin was administered for 3 weeks only and compared to amoxicillin/clavulanic acid21 as prospective trials. The authors identified that there is only a statistically significant difference favoring the macrolide therapy at one (of three different) time point (12 weeks after discontinuation of a 12-week medication) in the SNOT-scores (standard mean difference of 20.43 [20.82, 20.05]). All other investigated parameters failed to show significance. However, the clinical relevance of this difference in the SNOT score is doubtful. The present study is the third randomized, doubleblind, placebo-controlled trial on macrolides in chronic rhinosinusitis and the first after a surgical intervention. One multicenter study with 60 patients and a medication of azithromycin revealed no advantage over Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
Fig. 5. Results of the percentual changes in different parameters between visit 1 and visit 2 in the subgroups of CRSwNP and CRSsNP in the ERT and PBO groups. * indicates significant changes (P < 0.05). CRSsNP 5 cases without nasal polyposis; CRSwNP 5 cases with nasal polyps; ERT 5erythromycin; PBO 5 placebo; SNOT 5 Sino-Nasal Outcome Test; VAS 5 visual analogue scale; ENDO 5 endoscopy score. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
placebo.14 Only the parameter facial pain improved significantly in the azithromycin group. The DBPC monocenter Wallwork et al. trial, with 64 patients using roxithromycin over a period of 12 weeks, showed a higher reduction of the mean symptom response score compared to the placebo group, as well as a more pronounced and statistically significant change in the SNOT-20 score from baseline in the roxithromycin group at the end of treatment (0.46 points on a scale of 0 to 5).13 In this study, especially the subgroup of patients with low serum IgE showed significant improvements in the saccharine transit time and nasal endoscopic score and SNOT-20 score after 12 weeks, which could not be proven in our study. Similar to the results of the Wallwork et al. study13 with only CRSsNP patient included, in our study, the subgroup of CRSsNP patients also showed a tendency to a greater improvement after the macrolide therapy in most of the evaluated parameters (Fig. 5). It must be admitted that some methodological problems have occurred in our study: 1. The initially planned number of 70 participants could not be achieved, and the study was prematurely closed because the recruitment of patients could not be completed in a reasonable period of time. The study medication would have to be revolved after 3 years, and the costs for this action were too extensive in a nonfunded investigator-initiated trial. Therefore, the study unfortunately lost power. 2. The primary outcome criteria with concentrations of ECP and MPO in nasal secretions were initially chosen with the intention to receive objective results. In previous studies with long-term, low-dose clarithromycin in CRS patients, ECP as well as MPO levels in nasal secretions decreased
Laryngoscope 00: Month 2014
under the treatment,10 similar to a study with doxycycline treatment.22 These observations could not been reproduced in our trial; a reason might be the surgical trauma influencing the inflammatory markers substantially in our study. After surgical interventions, an inflammatory mucosal reaction was described, leading to an increase of inflammatory markers in nasal secretions. In patients suffering from cystic fibrosis who had surgery for CRS, ECP levels in nasal secretions stayed high or even increased after 1 month and declined only 4 months after surgery.23 The inflammatory reaction seems to outweigh possible beneficial effects of the macrolide treatment. Additionally, the volume of the nasal secretions gained by the sponge method that was used was too low, and results for MPO could only be gathered in a small amount of patients. Therefore, the initially planned determination of the concentrations of other specific markers (IL-8 and IL-5) also had to be abandoned completely. 3. The surgical intervention itself might have influenced the course of action so strongly that a possible additional effect of a macrolide therapy was covered. One study compared a medical treatment for CRS (with and without nasal polyps) with macrolides (loading dose of erythromycin 500 mg twice daily for 2 weeks, followed by 250 mg twice daily for 10 weeks) to a surgical intervention.24 Here, a significant increase in subjective and objective parameters in both groups could be found up to 1 year of follow-up without significant differences between the groups despite nasal volume, which was in favor for the surgical group. In a recently published review,25 additional controlled trials also in a postoperative setting are proposed.
Nevertheless, in our study with low-dose erythromycin 250 mg per day over a period of 3 months, the nasal endoscopy score showed a significant improvement in erythromycin-treated patients over placebo during the Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
7
study period. The amount of the difference of 1 point (out of 7 possible) also seems to be clinically relevant. In the follow-up period of an additional 3 months, this endoscopy score also decreased to lower values in the placebo group (Fig. 4). The other evaluated secondary parameters (including SNOT-20 score, self-rating of nasal health on a VAS, olfaction, and saccharin transit time) decreased in both groups over time and showed no statistically significant differences between the two treatment groups. Additionally, no positive effect on the primary criteria (ECP and MPO in nasal secretions) was found. In our study, no severe side effects were seen, and the study medication did not cause higher rates of severe adverse events compared to placebo. These results are in accordance with previous reports.13,26,27 But the ratio of side effects, especially of the gastrointestinal tract, was higher in the erythromycin group, which lead to a higher number of dropouts. Therefore, a higher dosage as used in other studies might cause even more side effects. It might well be possible that a macrolide medication over 12 weeks, as used in our study, is not long enough. As in another nonplacebo-controlled postsurgical study, the visual analogue scale scores for rhinorrhea and postnasal drip at 12 months after surgery were lower in a group that was treated with clarithromycin (200 mg/day) for 6 months than in the group with 3 months of treatment.28 One general problem in these types of studies is the limited number of patients who can be included. Both previously published DBPC studies showed similar patient numbers to the present study, one even in a multicenter design. Therefore, the results must always be discussed with caution because more included patients would mean more statistical power. On the other hand, investigator-initiated trials often are confronted with critical funding and duration is limited.
CONCLUSION In conclusion, the nasal markers for inflammation (ECP and MPO) could not be influenced by a long-term, low-dose erythromycin therapy in patients after sinus surgery. Despite a positive effect on the nasal endoscopy score in the treatment group, no other clinical advantage of the study medication over placebo could be proven. Therefore a general recommendation for the use of longterm, low-dose erythromycin therapy after surgery for CRS cannot be advocated. A presumably relevant improvement in the subgroup of patients without nasal polyps (CRSsNP) and specific endotypes (neutrophilic inflammation) should be further investigated.
BIBLIOGRAPHY 1. Fokkens WJ, Lund VJ, Mullol J, et al. EPOS 2012: European position paper on rhinosinusitis and nasal polyps 2012. A summary for otorhinolaryngologists. Rhinology 2012;50:1–12.
Laryngoscope 00: Month 2014
8
2. Bachert C, Zhang N, van Zele T, Gevaert P, Patou J, van Cauwenberge P. Staphylococcus aureus enterotoxins as immune stimulants in chronic rhinosinusitis. Clin Allergy Immunol 2007;20:163–175. 3. Gosepath J, Pogodsky T, Mann WJ. Characteristics of recurrent chronic rhinosinusitis after previous surgical therapy. Acta Otolaryngol 2008; 128:778–784. 4. Ponikau JU, Sherris DA, Kita H. The role of ubiquitous airborne fungi in chronic rhinosinusitis. Clin Allergy Immunol 2007;20:177–184. 5. Van Zele T, Claeys S, Gevaert P, et al. Differentiation of chronic sinus diseases by measurement of inflammatory mediators. Allergy 2006;61: 1280–1289. 6. Zhang N, Van Zele T, Perez-Novo C, et al. Different types of T-effector cells orchestrate mucosal inflammation in chronic sinus disease. J Allergy Clin Immunol 2008;122:961–968. 7. Smith TL, Litvack JR, Hwang PH, et al. Determinants of outcomes of sinus surgery: a multi-institutional prospective cohort study. Otolaryngol Head Neck Surg 2010;142:55–63. 8. Mendelsohn D, Jeremic G, Wright ED, Rotenberg BW. Revision rates after endoscopic sinus surgery: a recurrence analysis. Ann Otol Rhinol Laryngol 2011;120:162–166. 9. Wallwork B, Coman W. Chronic rhinosinusitis and eosinophils: do macrolides have an effect? Curr Opin Otolaryngol Head Neck Surg 2004;12: 14–17. 10. Cervin A, Wallwork B, Mackay-Sim A, Coman WB, Greiff L. Effects of long-term clarithromycin treatment on lavage-fluid markers of inflammation in chronic rhinosinusitis. Clin Physiol Funct Imaging 2009;29: 136–142. 11. Wallwork B, Coman W, Feron F, Mackay-Sim A, Cervin A. Clarithromycin and prednisolone inhibit cytokine production in chronic rhinosinusitis. Laryngoscope 2002;112:1827–1830. 12. Elmorsy S, El-Naggar MM, Abdel aal SM, Abou-elela MA. Sinus aspirates in chronic rhinosinusitis: fungal colonization of paranasal sinuses, evaluation of ICAM-1 and IL-8 and studying of immunological effect of longterm macrolide therapy. Rhinology 2010;48:312–317. 13. Wallwork B, Coman W, Mackay-Sim A, Greiff L, Cervin A. A double-blind, randomized, placebo-controlled trial of macrolide in the treatment of chronic rhinosinusitis. Laryngoscope 2006;116:189–193. 14. Videler WJ, Badia L, Harvey RJ, et al. Lack of efficacy of long-term, lowdose azithromycin in chronic rhinosinusitis: a randomized controlled trial. Allergy 2011;66:1457–1468. 15. Fokkens W, Lund V, Mullol J. European position paper on rhinosinusitis and nasal polyps 2007. Rhinol Suppl 2007:1–136. 16. Piccirillo JF, Merritt MG Jr, Richards ML. Psychometric and clinimetric validity of the 20-item sino-nasal outcome test (SNOT-20). Otolaryngol Head Neck Surg 2002;126:41–47. 17. Baumann I. [Validated instruments to measure quality of life in patients with chronic rhinosinusitis]. HNO 2009;57:873–881. 18. Wolfensberger M, Schnieper I, Welge-Lussen A. Sniffin’Sticks: a new olfactory test battery. Acta Otolaryngol 2000;120:303–306. 19. Fokkens W, Lund V, Mullol J. EP3OS 2007: European position paper on rhinosinusitis and nasal polyps 2007. A summary for otorhinolaryngologists. Rhinology 2007;45:97–101. 20. Pynnonen MA, Venkatraman G, Davis GE. Macrolide therapy for chronic rhinosinusitis: a meta-analysis. Otolaryngol Head Neck Surg 2013;148: 366–373. 21. Amini M, Yarmohammadi M, Izadi P. A comparing study of clarithromycin XL with co-amoxiclav for treatment of chronic sinusitis; a clinical trial. Iranian Journal of Clinical Infectious Diseases 2009;4:197–201. 22. Van Zele T, Gevaert P, Holtappels G, et al. Oral steroids and doxycycline: two different approaches to treat nasal polyps. J Allergy Clin Immunol 2010;125:1069–1076 e1064. 23. Weber A, Kiefer J, Peters S, Schneider M, Bargon J, May A. Eosinophilic cationic protein as a marker of nasal inflammation in patients with cystic fibrosis. Laryngoscope 1999;109:1696–1702. 24. Ragab SM, Lund VJ, Scadding G. Evaluation of the medical and surgical treatment of chronic rhinosinusitis: a prospective, randomized, controlled trial. Laryngoscope 2004;114:923–930. 25. Cervin A, Wallwork B. Efficacy and safety of long-term antibiotics (macrolides) for the treatment of chronic rhinosinusitis. Curr Allergy Asthma Rep 2014;14:416. 26. Cervin A, Wallwork B. Macrolide therapy of chronic rhinosinusitis. Rhinology 2007;45:259–267. 27. Videler WJ, van Hee K, Reinartz SM, Georgalas C, van der Meulen FW, Fokkens WJ. Long-term low-dose antibiotics in recalcitrant chronic rhinosinusitis: a retrospective analysis. Rhinology 2012;50:45–55. 28. Nakamura Y, Suzuki M, Yokota M, et al. Optimal duration of macrolide treatment for chronic sinusitis after endoscopic sinus surgery. Auris Nasus Larynx 2013;40:366–372.
Haxel et al.: Long-Term Low-Dose Erythromycin and CRS
