Benefits and harm of systemic steroids for short- and long-term use in rhinitis and rhinosinusitis: an EAACI position paper

1. Allergic rhinitis

AR is the most prevalent presentation form of all allergic diseases and the most com-mon chronic disorder in children. It is considered a risk factor for the development of asthma and a major public health problem, due to its prevalence and impact on patients’ quality of life, work/school performance, and economic burden [30. ].

Intranasal GCS and oral/topical antihistamines are the most effective symptomatic treatment for AR and should be the first-line therapy for mild to moderate disease [30., 31. ]. Moderate to severe disease not responsive to intranasal GCS, should be treated with additional pharmacological therapies (including cromolyns and leukotriene receptor antagonists), allergen immunotherapy (AIT) and non-pharmacologic therapies (such as nasal irrigation) [30., 31. ]. Usually a combination of intranasal GCS and a topical or oral antihistamine is used for moderate to severe AR.

Regarding the use of systemic GCS in AR, the current evidence is scarce. Three studies compared the effect of systemic GCS in adult patients (> 15 year old) with AR (Table 2 ).

The first randomized controlled trial (RCT) from 1987 showed a beneficial effect of a depot injection of 80 mg methylprednisolone (MP) vs. placebo on nasal obstruction and eye symptoms in 48 AR patients, which lasted for 4 weeks [32. ]. The second study by Brooks et al. [33. ] investigated the efficacy of different doses of oral MP and placebo in patients not treated with other medications. Thirty-one patients were randomized to receive 0, 6, 12, or 24 mg MP. Oral GCS produced dose-related reduction in all symptoms. The difference between placebo and 24 mg MP was significant for all the symptoms monitored, except itching, which benefited marginally. With 6 mg MP, congestion, drainage, and eye symptoms showed significant drug-placebo differences, but itching, running/blowing, and sneezing did not. The third study by Laursen et al. [34. ] compared prednisone 7.5 mg for 3 weeks with a single intramuscular injection of betamethasone dipropionate also in patients not treated with other medications. This study showed a therapeutic index in favour of the depot injection versus oral treatment in AR [33. ].

2. Non-allergic rhinitis

Although, the prevalence of NAR among the chronic rhinitis patients ranges from 20 to 50% [36. ], their disease mechanisms and treatment options are much less studied than their allergic peers. NAR comprises a heterogeneous group of chronic rhinitis subtypes, such as drug-induced rhinitis, hormonal-induced rhinitis, some forms of occupational rhinitis and rhinitis linked to systemic diseases [37. ]. However, in about 50% of the NAR patients, no specific causal factor can be found and this is addressed as idiopathic rhinitis (IR) [37. ]. Up till now, no studies are available that investigate the effectiveness of systemic steroids in NAR or IR patients. However, since it is believed that in IR neurogenic pathways are involved, rather than classical inflammatory pathways [38. ], systemic GCS are not the therapy of choice. Of note, all IR patients included in a recent study investigating the effect of capsaicin in IR, reported lack of clinical response to intranasal GCS [38. ]. By extrapolation, there is a low likelihood of oral GCS being effective in this patient population, unless more than one etiologic or inflammatory mechanism underlies the development of rhinitis.

3. Acute rhinosinusitis

Compared to the literature on effectiveness of systemic GCS in CRS, data on acute rhinosinusitis (ARS) are scarce. In 2014 an update of a Cochrane review was published [40. ] concluding that systemic GCS as a monotherapy are ineffective compared to placebo in ARS patients, but might have a beneficial effect on short-term symptom relief when used as an adjunctive therapy to antibiotics.

Up to date, five randomized, placebo-controlled trials investigating the effect of oral GCS in adults with ARS are available and included in the Cochrane meta-analysis (Table 3 ). From those, only one focused on systemic GCS as a monotherapy [41. ]. In this high-quality second-line clinical trial, patients with clinically diagnosed ARS were randomized to receive either prednisolone 30 mg/day or placebo for 7 days. In the 174 patients who completed the trial, no clinically relevant benefit of prednisolone over placebo was found regarding facial pain or pressure, other nasal symptoms or quality of life.

Four other RCTs investigated the adjunctive effect of systemic GCS to oral antibiotics in ARS. Gehanno et al. [42. ] reported the adjuvant effect of 5 days of 3 × 8 mg MP/day to amoxicillin–clavulanate in 417 patients. On day four, patients showed significantly less pain in the steroid group whereas nasal discharge did not significantly improve. The use of additional medication was not reported.

In 2004, two similar studies were published; a French study [43. ] showed a beneficial effect on pain with oral prednisone as an add-on therapy to cefpodoxime in 291 ARS patients. Also Ratau et al. [44. ] reported a significant benefit of 1 mg of oral betamethasone per day as adjunct to amoxicillin–clavulanate in 42 patients.

In 1990 Cannoni already published similar findings showing a better symptom resolution in ARS patients treated with 40 mg prednisolone/day in combination with antibiotics, compared to patients receiving a non-steroidal anti-inflammatory drug (NSAID) with antibiotics [45. ].

4. Chronic rhinosinusitis without nasal polyps

For clinical purposes, the definition of CRS includes nasal polyposis (NP) and currently it is still unclear why some CRS patients develop NP and others do not. CRSsNP is characterized by basement membrane thickening, goblet cell hyperplasia, fibrosis, subepithelial oedema and influx of inflammatory cells that are mainly of the neutrophilic subtype with a cytokine pattern deviated towards the Th1 subtype [5. ].

Based on available data, medical therapy for CRS should begin with daily application of intranasal steroids in conjunction with saline irrigation and subsequent therapies are based on the patient's severity of symptoms and/or quality of life impairment [4. ].

There is limited data showing efficacy of oral GCS in CRSsNP and a systematic review analysed the available literature in 2011 [46. ].

No RCT investigated the effects of oral GCS in CRSsNP and only two retrospective case series in adults are available [47., 48. ] that both considered CRSwNP and CRSsNP patients, but sub-group analysis allowed an evaluation specific to CRSsNP (Table 4 ). Both retrospective studies investigated the effects of oral prednisone in conjunction with 1 month of oral antibiotics added to intranasal steroids and irrigations. Improved subjective and objective outcomes were seen after multimodality treatment schemes in both studies for CRSsNP. The study of Subramamian et al. [48. ] pooled both CRSwNP and CRSsNP patients and found that the CRSsNP patients had better outcomes than CRSwNP patients. Lal et al. [47. ] demonstrated that the CRSsNP patients showed total symptom resolution 2 months after treatment of 54.9% compared to 51% for the total CRS group. There are no studies available that investigated the benefits of systemic GCS in monotherapy in treating CRSsNP.

5. Chronic rhinosinusitis with nasal polyps

CRSwNP is different from CRSsNP by the presence of nasal polyps consisting of a large quantity of extracellular oedema with the presence of a dense inflammatory cell infiltrate [49., 50. ], which is characterized in about 80% of the Caucasian CRSwNP patients, by activated eosinophils [51., 52. ] and is associated with a predominant Th2 cytokine profile (IL-4, IL-5, IL-10, eotaxin) [53., 54. ].

A recent suite of Cochrane Reviews has considered the efficacy of interventions for CRSwNP. Two reviews were performed with respect to short-term oral GCS; one comparing oral GCS alone versus placebo or other treatment [55. ], and a second comparing oral GCS used as an adjunct to other treatments, versus control [56. ].

For oral GCS alone, 8 trials with a total of 474 participants, all of whom were adult patients CRSwNP, were identified [57.-64. ]. All studies followed up patients to the end of the treatment course, and 3 followed patients for 3 to 6 months after completion. Patients receiving oral GCS achieved better quality of life (standardized mean difference (SMD) of − 1.24 95% CI − 1.92 to − 0.56, measured with RSOM-31), lower nasal symptom scores (SMD − 2.84, 95% CI − 4.09 to − 1.59) and greater polyp reduction (SMD − 1.21) than control groups at the end of the course of treatment. However, there was no difference between groups at 3 to 6 months after the course of treatment.

Treatment doses utilized in included studies included prednisone at 30 mg and reduced over 14 days, prednisolone at 60 mg reducing over 17 days, or at constant dosage of 50 mg or 25 mg for 14 days, or reducing dosages of MP over 20 days. Of the three studies that followed patients beyond the course of treatment, 2 prescribed ongoing intranasal GCS after completion of the systemic dose to both groups while one did not [58., 62., 63. ].

Included trials were considered to be at low risk of bias, but overall the quality of evidence was rated as low due to the small numbers of participants, heterogeneity of outcome measures and limited follow-up time in most studies.

Another trial considered oral GCS versus placebo as an adjunct to treatment with intranasal GCS in CRSwNP patients [65. ]. This study recruited 30 participants and was considered at high risk of bias because of lack of blinding and lack of information on randomization. It reported greater reduction in polyp size in the active treatment arm (MD − 0.46, 95% CI − 0.87 to − 0.05).

One trial included in the Cochrane review of oral GCS as an adjunctive treatment recruited children [66. ] and is therefore considered later in this document.

6. Allergic fungal rhinosinusitis

7. Nasal manifestations of auto-immune disease

Many auto-immune disorders can involve the nose: thyroid auto-immunity, various vasculitis, Sjogren's syndrome and sarcoidosis are the most frequently encountered, but other connective tissue diseases, such as systemic lupus erythematosus, polyarteritis nodosa, scleroderma and relapsing polychondritis can also have nasal symptoms [39. ].

GCS have been the major therapeutic option for some of these diseases as an immune suppressant for the past decades, probably being most effective where eosinophils, which are exquisitely steroid-sensitive, are involved [79. ]. However, the quality of the evidence for their efficacy is poor, with studies mostly being reviews or open pilots, even in seminal trials such as those of Fauci for Wegener's granulomatosis [80.-82. ]. The reasons for this include not only time-hallowed use, but also difficulty in undertaking placebo-controlled trials in severe diseases, differences in the manifestations and their intensity between individual patients, disease complexity and plasticity and probably lack of interest in funding. This situation is now changing with the advent of newer therapies, particularly monoclonal antibodies, which are being trialled against older therapies including GCS [83. ].

Churg–Strauss syndrome, now called eosinophilic granulomatosis with polyangiitis (EGPA), is classically considered a Th2-mediated disease and affects sino-nasal mucosa in > 80% of the patients. Treatment must be tailored according to prognostic factors identified by the French Vasculitis Study Group [84. ]. GCS alone are used for mild disease, high-dose GCS and cyclophosphamide is still the gold standard for severe cases [85. ], but biological agents such as rituximab or anti-IL-5 biologicals are promising, though costly, alternatives [86. ].

The hallmark of granulomatosis with polyangiitis (GPA; previously known as Wegener's disease) is the coexistence of vasculitis and granuloma and again over 80% of patients show sino-nasal involvement [87. ]. GCS alone are insufficiently effective: the induction treatment for severe GPA comprises GCS combined with another immunosuppressant, cyclophosphamide or rituximab. Once remission is achieved, maintenance strategy following cyclophosphamide-based induction relies on less toxic agents such as azathioprine or methotrexate.

GCS decrease the frequency, duration, and severity of flares in relapsing polychondritis, but do not stop disease progression in severe cases [88. ].

The presence of sino-nasal disease is associated with more severe sarcoidosis and the need for systemic GCS therapy [89. ].

Treatment for systemic lupus erythematosus (SLE) by various organ systems is not evidence-based beyond the usual first- or second-line treatment, however a recent meeting achieved consensus in several scenarios, including anti-phospholipid syndrome [90. ].

GCS, often combined with NSAIDs, are used in Sjogren's syndrome to treat associated interstitial lung disease and/or sensorineural hearing loss [91. ].

8. Sino-nasal pathology and concomitant asthma

Asthma is a chronic inflammatory disease of the lower airways involving inflammation of the bronchial mucosa, and variable obstruction of bronchi due to intrinsic/extrinsic stimuli, and leading to symptoms such as episodic breathlessness and wheezing with airway hyperresponsiveness to environmental stimuli [92. ]. Since the introduction of the “United Airway Disease” concept [1. ], a large series of scientific publications from clinical epidemiology, pathophysiology, histology, and treatment outcomes has correlated asthma and upper airway disease. AR and asthma often coexist and AR is regarded as a risk factor for the development of asthma. Uncontrolled rhinitis impacts asthma control. Asthmatic patients have a higher CRS severity score than non-asthmatic patients, and more nasal polyps, indicative of a strong relationship between CRS severity and asthma [93. ]. It has been reported that 20–60% of patients with CRSwNP have asthma [94., 95. ].

The first use of GCS to treat acute asthma exacerbation was in 1956 [96. ]. Development of GCS that have less mineralocorticoid activity, like prednisone, and later those that have no mineralocorticoid activity, like dexamethasone, made steroid use more attractive therapies to use in asthma. Prescribing a short course of oral GCS following the treatment of acute asthma exacerbations was found to reduce the rate of relapse [97. ]. However, courses longer than 5 days were not found to provide any additional benefit [98. ].

As described above, systemic GCS should not be considered as a treatment for AR. We could not identify any systematic review, randomized trial, or controlled study that evaluated the use of systemic GCS in patients with AR with concomitant asthma not responding to other therapy.

When analysing the evidence of oral GCS for patients with CRS and coexisting asthma there are a few randomized controlled trials and uncontrolled prospective interventional studies that evaluated the efficacy of different treatments (Table 9 ) of which only one looked at systemic GCS use. This study was carried out in adults by Ikeda et al. [99. ] and included 21 CRS patients with concomitant asthma. Fifteen patients underwent ESS, and 6 other patients remained on medical therapy. Seven patients of the ESS group showed a reduction in the need for GCS during the 6 months following surgery, whereas two patients were unchanged and two patients required larger dosages.

1. Hypothalamic–pituitary–adrenal-axis (HPA) inhibition

Reductions in the level of plasma cortisol are reported after one injection of GCS. They usually decrease in the first 2 weeks after steroid administration, but slowly return to normal after 3 weeks, as has been demonstrated in patients with AR [103. ]. Hedner et al. [104. ] showed a minor HPA dysfunction in 14 allergic patients treated with a single intra-muscular injection of MP acetate, which returned completely to normal at 4 weeks post-injection. In a double-blind study by Laursen et al. [105. ] 36 birch pollen allergic patients were treated with either a single injection of betamethasone dipropionate or oral prednisolone 7.5 mg/day for 3 weeks. Only the prednisolone treated patients showed reduction in plasma cortisol levels at 3 weeks.

Bonfils et al. [106. ] prospectively evaluated the HPA-axis in patients with CRSwNP (n = 46), who received at least three short courses of oral GCS in the last year (course 6–8 days, 1 mg/kg/day, mean duration of treatment 4.7 years, mean 6.8 courses/year, mean cumulative prednisone consumption 3,800 mg). The study demonstrated that 48% of patients had an asymptomatic adrenal insufficiency diagnosed with the Synacthen test.

2. Hyperglycemia and diabetes

A retrospective study based on Danish National Registries, including 47,382 AR patients, demonstrated that treatment with at least one consecutive injection of depot corticosteroid for 3 years on a row was associated with an increased risk of being diagnosed with diabetes later in life (RR 1.4) [107. ]. The degree of new-onset diabetes associated with intermittent short-term oral GCS has not been clearly established.

3. Osteoporosis

In the same Danish epidemiological study, Aasbjerg et al. [107. ] showed that, compared to immunotherapy, treating AR with annual depot-steroid injections (i.e. at least one steroid injection in the pollen season for 3 consecutive years) was associated with increased risk of being diagnosed with osteoporosis (RR 1.2). The above-mentioned study from Bonfils, investigating the HPA-axis, prospectively evaluated the occurrence of osteoporosis in patients with CRSwNP (n = 46), receiving at least three short courses of oral GCS in the previous year. Osteopenia of the proximal femur was present in 40.5% and osteoporosis was present in 54% [106. ]. Rajeskaran et al. [108. ] retrospectively evaluated the risk of osteoporosis in patients with CRS (n = 176), who received oral GCS ≥ 5 mg daily for 3 consecutive months any time in the past. Overall, low bone mineral densities (BMD; osteopenia or osteoporosis) was 38.6%. These studies were recently evaluated in a systematic review which was unfortunately not able to quantify the overall risk of osteoporosis induced by oral GCS for CRSwNP, due to the low number of studies [109. ].

The effects of short-course oral GCS on bone mineral density (BMD) have also been investigated in a 4-year longitudinal small study in asthmatic patients. Asthmatic patients receiving frequent short courses of oral GCS (i.e. > 2.5 courses/year; n = 9) compared to those receiving sporadic courses (i.e. ≤ 2.5 courses/year; n = 26) revealed a greater loss of lumbar BMD (T-score 82.0% versus T-score 77.7%) in the frequently treated group [110. ]. Also, a lower Z-score of 93.1% was demonstrated in frequent short courses, versus the sporadic courses that did not show a lower Z-score than the normal population values (Z-score 100.1%).

4. Avascular necrosis

With regards to avascular necrosis of the femoral head in patients treated with systemic GCS for upper airway disease, we found 1 case report of Nasser et al. [111. ] describing a single case with severe hay fever that was given at least one depot corticosteroid injection each year for 11 years, leading to avascular necrosis.

More individual case reports highlight the relationship between the use of systemic GCS and avascular necrosis. The risk to develop osteonecrosis seems to be dependent on the prescribed dose, the cumulative dose and route of administration, as well as underlying disease states (SLE patients seem to be particularly at risk) [112.-114. ].

5. Gastrointestinal disturbances and peptic ulceration

In a randomized double-blind placebo-controlled study by Kirtsreesakul et al. [62. ] 112 patients with CRSwNP used either 50 mg prednisone or placebo for 14 days and reported significantly more (mild) gastrointestinal disturbances and dyspepsia in the prednisolone treated group. In a double-blind placebo-controlled trial by Venekamp et al. [41. ] 174 adult patients clinically diagnosed with ARS received either 30 mg/day prednisolone or placebo for 7 days. The incidence of gastrointestinal complaints did not differ between treatment groups.

In a large nested case–control analysis based on the UK General Practice Research Database, 2105 cases of upper gastro-intestinal complications were compared to 11,500 controls and then evaluated for exposure to certain drugs e.g. corticosteroid use. The adjusted OR for current use of oral GCS was 1.8 (95% CI 1.3–2.4) for upper gastrointestinal complications overall [115. ]. No statistically significant difference could be objectified for lower versus higher dosage of GCS. To our knowledge no studies in upper airway disease patients report on systemic steroid treatment and peptic ulceration.

6. Ocular adverse effects

GCS have been described to induce the formation of posterior subcapsular cataract or glaucoma. The risk for patients using repeated (short) courses of systemic GCS for upper airway disease is currently unknown.

There is evidence in rheumatoid arthritis patients that this risk is enhanced after therapy lasting more than 1 year [116. ]. Another study by Huscher et al. [101. ] analysed dose-related patterns of self-reported symptoms from 1066 patients with RA with ongoing long-term (> 6 months) systemic GCS. These symptom patterns were compared to non-users (no systemic GCS for at least 12 months). The prevalence of self-reported cataract was higher for all dosages of GCS, whereas the prevalence of self-reported glaucoma was only increased in those taking > 7.5 mg/day (6.6% users vs. 2.7% non-users).

7. Infections

A meta-analysis of randomised controlled clinical trials in which patients were randomised to treatment with or without systemic GCS (n = 4198) showed that the rate of infection was not significantly increased in patients who were given a mean dose of less than 10 mg/day of prednisone or a cumulative dose of less than 700 mg [117. ]. This meta-analysis included a wide variety of diseases warranting systemic GCS. The true risk of developing infection in patients using short courses for upper airway disease remains uncertain.

8. Local adverse effects of steroid-injections

We found one case report on gluteal subcutaneous atrophy that was seen after a depot steroid injection of triamcinolone for AR [118. ]. A study of Laursen et al. [34. ] investigated specifically the reporting of all AE's related to GCS injections for AR to the ‘Danish Register for the Side-Effects of Drugs’ and evaluated the reported events consecutively for a 10-year period. The study demonstrated that one out of 11,785 injections came with any local AE. Most AE's were reversible and primarily skin related, such as skin atrophy.

9. Cardiovascular adverse effects

Cardiovascular disease is mainly associated with high dose and long-term use, primarily hypertension and acute myocardial infarction are described [100., 119. ].

A population-based cohort study in 68,781 GCS users and 82,202 non-users showed that patients exposed to dosages of GCS > 7.5 mg of prednisolone/day (or equivalent) during 1 to 5 years of follow-up, had substantially higher rates of myocardial infarction, heart failure, or cerebrovascular disease (adjusted RR of 2.56; 95% CI 2.18–2.99). The risk was not increased in patients using < 7.5 mg prednisolone equivalent daily [120. ].

Another large, retrospective case–control study with data extracted from the General Practice Research Database (1988–1997) showed in over 100,000 individuals that the use of oral GCS comes with a 25% higher risk of any cardiovascular or cerebrovascular outcome compared to controls. Current use (in the 3 months before the registration of an event) and highest average daily dose give a much stronger association. Current use is also associated with a significantly increased risk of heart failure (adjusted OR of 2.66; 95% CI 2.46–2.87) and ischemic heart disease (OR of 1.20; 95% CI 1.11–1.29), but not ischemic stroke or transient ischemic attack. Cardiovascular risk showed a clear dose–response relationship [121. ].

To our knowledge, the risk in patients using GCS for intermittent short courses is unknown.

10. Neuropsychiatric effects

A study from Hissaria et al. [60. ] investigating 40 CRSwNP patients treated with 50 mg of prednisolone daily for 14 days or placebo, found that sleep disturbances were reported as a significant prevalent AE (40%) compared to placebo (10%). Mood disturbance were more frequently reported, but not significantly different from placebo (25% vs. 10%).

In the above-mentioned controlled trial by Venekamp et al. [41. ] studying ARS patients treated with 30 mg/day prednisolone or placebo for 7 days, the incidence of mood or sleep disturbance did not differ between treatment groups.

Two studies in asthmatic and ophthalmologic patients receiving short-courses of GCS, showed a development of (hypo)mania [122., 123. ] as well as depression symptoms [123. ].

Naber et al. [123. ] showed in a prospective uncontrolled study in ophthalmologic patients receiving systemic GCS (n = 50) that 26–34% of patients developed (hypo)mania and 10–12% developed depression syndromes when using an initial 119 ± 41 mg/day MP or fluorcortolone, tapered to 75 ± 22 mg/day at 8 days. The onset of symptoms was within 3 days of use and there was no correlation between daily dose and daily ratings of mood. Brown et al. [122. ] showed in 32 asthmatic patients using prednisone (mean course 13.9 days, mean dose of 36.9 mg/day) a highly significant increase in self-reported mania, but no increase in depression during the first 3–7 days of therapy. Mood changes returned back to normal after discontinuation of therapy.

11. Cushingoid features

We found no studies investigating Cushingoid appearance in rhinitis/rhinosinusitis patients treated with GCS and only a few studies addressed the risk of intermittent short courses of GCS and weight gain.

A randomised controlled trial by Campieri et al. [124. ] in patients with active Crohn's disease demonstrated that 38% of patients on a regimen of prednisolone tapered over 12 weeks (40–45 mg) developed a ‘moon face’. Mean body weight increased with 2.1 kg after 8 weeks of treatment. Bar-Meir et al. [125. ] showed that patients receiving 8 weeks of prednisone developed a moon face in 33% versus 16% in patients receiving a similar treatment with budesonide.

1. Efficacy of systemic GCS in pediatric CRS and ARS

Three clinical trials can be found in literature that investigated the use of oral GCS in the pediatric rhinosinusitis population, of which only one is controlled (Table 10 ).

This controlled study involved 48 children (mean age 8 years) with CRSsNP [66. ] and investigated the effect of oral GCS as an add-on to antibiotics. 22 participants received either 30-day course of oral amoxicillin–clavulanate and 15-day course of oral MP and 23 participants received only antibiotics and a placebo. The mean change of total symptom score and CT score was significantly higher after treatment with oral GCS and antibiotics compared with placebo and antibiotics (P < 0.001). There was also a significant beneficial effect of oral GCS in cough, nasal obstruction and post-nasal drainage symptom scores. Complete clinical recovery after 30 days of treatment was obtained in significantly more subjects receiving MP (P < 0.005). Recurrence of symptoms 6 months after the end of treatment was not statistically significant between the groups.

Additionally, a retrospective study involving 35 young CRS patients (1–21 years) undergoing serial sinus CT scans due to medical reasons, evaluated Lund Mackay ostiomeatal complex score in relation to three different treatment schemes [127. ] antibiotics, intranasal topical GCS and oral systemic GCS. The data suggested that the use of systemic GCS was associated with a significant increase in the likelihood of radiologic improvement. The retrospective study design, the small and heterogeneous population, heterogeneous treatment modalities, and the lack of adjustments, limit the possibilities to assess clinical significance of the findings.

A second uncontrolled study [5. ] evaluated cytokine pattern of 30 asthmatic CRS patients (4–12 years) before and after the treatment of amoxicillin–clavulanate, fluticasone propionate aqueous nasal spray and a short course of oral deflazacort. After the treatment, endoscopic resolving of mucopurulent discharge was detected in 25/30 children, the median concentration of IL-4 decreased significantly in all subjects, and the median IFN-γ concentration increased significantly only in the atopic subgroup (N = 16). The uncontrolled study design and uncertainty whether the patients used prescribed drugs, limits the possibilities to assess effect of systemic GCS.

2. Harm of GCS in children

There is limited knowledge of risks of using systemic GCS in pediatric CRS or ARS compared to pediatric asthma. As an example, the Childhood Asthma Management Program trial followed the annual bone mineral accretion of 877 children (5–12 years) with mild-to-moderate asthma [128., 129. ]. Oral GCS bursts produced a dosage-dependent reduction in bone mineral accretion (0.052, 0.049, and 0.046 g/cm² per year) and an increase in risk for osteopenia (10%, 14%, and 21%) for 0, 1–4, and ≥ 5 courses, respectively, in boys. The authors conclude that multiple oral GCS bursts over a period of years can produce a dosage-dependent reduction in bone mineral accretion and increased risk for osteopenia in children with asthma. 780 children with asthma were followed for a mean of 4.3 years and it was shown that boys with lower vitamin D levels are significantly more susceptible to the negative effects of GCS on bone mineral accretion over time [129. ]. Regarding studies investigating GCS AE's in upper airway disease, the trial from Ozturk also looked at self-reported AE's during the 15-day course of oral MP [66. ]. In this trial no clinically significant AE's were reported. At the end of the treatment, the mean weight change did not differ statistically significantly between the groups. No data of monitored AE's, nor that of long-term outcomes, nor that of bacterial culture were available in this study.

A systematic review has been performed to determine the most common and serious drug-related AE of long courses of oral GCS in children [130. ]. Literature search of several databases was performed to identify all studies in which systemic GCS had been administered to pediatric patients ranging from 28 days to 18 years of age for at least 15 days of treatment. The group found 91 studies that represented a total of 6653 children and contained reports of 4124 adverse drug reactions, the majority in patients with leukaemia, haemangioma and asthma. The three most frequent adverse drug reactions were weight gain (22.4%), Cushingoid features (20.6%) and growth retardation (18.9%). Increased susceptibility to infection was the most serious adverse drug reaction. 24 children died from infections, 10 from varicella zoster.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

GS: Honoraria for articles, speaker and advisory boards: ALK, Astra Zeneca, Brittania Pharmaceuticals, Capnia, Church & Dwight, Circassia, Groupo Uriach, GSK, Meda/Mylan, Merck, MSD, Ono Pharmaceuticals, Oxford Therapeutics, Sanofi-Aventis, UCB. Travel funding: ALK, Bayer, GSK, Meda. Chair of BSACI Rhinitis guidelines, EAACI Ethics Committee, Rhinology & Laryngology Research Fund. Chair of Data Monitoring Board for Acarizax paediatric AR trial.

CH: Reimbursed for Advisory Board work for sanofi, Smith and Nephew and speakers bureau for Medtronic.

CB: Advisory board of Sanofi, GSK, Novartis, Astra-Zeneca, Mylan and reimbursed for presentations and travel.

TVZ: Consultant for Medtronic and 3NT.

STS: has acted as paid consultant for ERT and Roche Products. All these are outside the submitted work.

All other authors declare that they have no competing interests.