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

Chronic upper airway inflammation is one of the most prevalent chronic disease entities in the world with rhinitis being the most common presentation form affecting 30% of the Western population [1].

Rhinitis is defined as an inflammation of the lining of the nose and is characterized by nasal symptoms including rhinorrhoea, sneezing, nasal blockage and/or itching of the nose. Allergic rhinitis (AR) is the best-known form of non-infectious rhinitis and is associated with an IgE-mediated immune response against allergens [1]. However, a substantial group of rhinitis patients has no known allergy and they form a very heterogeneous non-allergic rhinitis (NAR) patient population suffering from drug-induced rhinitis, occupational rhinitis, irritant-induced rhinitis, hormonally linked rhinitis and idiopathic rhinitis [2, 3]. When inflammation of the nasal mucosa extends to the mucosa of the paranasal sinuses, the consensus term of rhinosinusitis is used. Rhinosinusitis has been shown to affect about 10% of the Western population [4]. In addition to rhinitis symptoms, rhinosinusitis is characterized by postnasal drip, facial pressure and reduction or loss of smell [5]. Acute rhinosinusitis (ARS) is a very common condition and mostly of viral origin [5]. About 0.5–2% of the viral ARS are complicated by a bacterial infection [5].

Chronic rhinosinusitis (CRS) is defined as the presence of two or more nasal symptoms, one of which should be either nasal blockage or nasal discharge, and/or smell problems, and/or facial pain for more than 12 weeks, in combination with inflammatory signs confirmed by nasal endoscopy and/or CT scan. CRS can either present with nasal polyps (CRSwNP) or without (CRSsNP). Additionally, chronic upper airway disease often coexists with lower airway problems, most frequently asthma, but also a link with chronic obstructive pulmonary disease (COPD) and bronchiectasis has been reported [6].

Glucocorticosteroids (GCS) are the oldest and most widely used anti-inflammatory therapy. Since their introduction in the 1950s, GCS have played a key role in the treatment of various inflammatory, allergic, and immunologic disorders. Consequently, they are known as a very effective drug for treating chronic airway inflammatory diseases involving both lower as well as upper airways [1, 4, 7]. GCS can be administered topical or systemically. If possible topical GCS are preferred over systemic GCS treatment as it is well known that this systemic GCS treatment is linked to an extensive range of potential adverse effects (AE’s) that have been well-described in the literature and vary from uncomfortable to life-threatening [8]. Notably, reports on AE and/or toxicity of systemic GCS cover a heterogeneous group of GCS-treated diseases, which complicates the interpretation of the actual risk for the rhinitis/rhinosinusitis patients.

Therefore, the risk–benefit ratio of treating non-life-threatening upper airway diseases with systemic GCS remains debatable and needs clarification.

This document summarizes the current evidence for beneficial as well as harmful effects of administration of systemic GCS in the different types of upper airway disease and aims at providing recommendations about its use in rhinitis and rhinosinusitis based on the current evidence. For each topic 2 experts in the field were appointed to review the literature and topics that were appropriate for clinical recommendations were considered as evidence-based reviews with recommendations. The experts then provided a recommendation based upon the guidelines of the American Academy of Pediatrics (following the recommendation strategy used by the International Consensus on Allergy and Rhinology [9]). Table 1 summarizes the recommendation development based on the combination between levels of evidence and the benefit/harm balance. Generally, the search was focused on adults. Two experts reviewed the literature specifically for the pediatric population.

The search was performed in the MEDLINE (Ovid 1946—current; and PubMed 1966—current) and Cochrane databases. The search strategy was based on a combination of MeSH-terms and free text words. Search terms are listed in Additional file 1.

Corticosteroids, which are produced by the adrenal glands, can be classified as glucocorticoids and mineralocorticoids. Cortisol is the endogenous glucocorticoid in humans, naturally derived from cholesterol metabolism upon stimulation by the hypothalamic–pituitary–adrenal axis (Fig. 1), which is regulated initially by the circadian rhythm, but also by negative feedback by glucocorticoids and glucocorticoid increment induced by stressors such as pain, inflammation or infections [10].

GCS are involved in several physiologic functions. They control the metabolism of carbohydrates, proteins and lipids, as well as the balance of calcium [11, 12]. However, the most explored effects of GCS are the anti-inflammatory and immune-suppressive functions. GCS inhibit the activation and survival of inflammatory cells and modulate the activity of structural cells [13, 14]. The main anti-inflammatory effects of GCS are based on their ability to reduce the synthesis of several cytokines (IL-1, -2, -3, -4, -5, -6, -8, TNF-α, IFN-γ, GM-CSF) from many cells (macrophages, monocytes, lymphocytes, fibroblasts, and epithelial and endothelial cells). This affects recruitment, localization, protein synthesis, and survival of inflammatory cells such as eosinophils [15]. The recruitment of inflammatory cells is also diminished by an inhibited expression of adhesion molecules such as ICAM-1 and VCAM-1 [16], which affects the influx of basophils and mast cells in the epithelial layers of nasal mucosa. Finally, GCS are involved in the pathological wound repair mechanism called remodelling. Remodelled tissue such as the stroma of nasal polyps contains abundant infiltration of inflammatory cells, increased fibroblasts numbers and increased extra-cellular matrix deposition. However, GCS appear to be minimally effective in reversing the structural changes resulting from remodelling [17].

All these effects are exerted by intracellular activation of the glucocorticoid receptor (GR) [18]. The GR belongs to the superfamily of ligand regulated nuclear receptors [19] and alternative splicing of the GR primary transcript generates two receptor isoforms, named GRα and GRβ. GRα has a widespread distribution in cells and tissues [20], including healthy and diseased upper airway mucosa. Inactive GRα is found primarily in the cytoplasm of cells as part of a large multi-protein complex [21]. Glucocorticoids diffuse across the cell membrane and bind to GRα resulting in a nuclear entry (Fig. 2) [22] where GRβ modulates either positively or negatively the expression of target genes. GRβ has a very low level of expression compared to GRα [20] and acts mainly as a negative inhibitor of GRα-mediated gene modulation [23].

The anti-inflammatory effects of GCS are explained by three broad molecular mechanisms: the decreased expression of pro-inflammatory genes (trans-repression), the increased expression of anti-inflammatory genes (trans-activation), and non-genomic mechanisms. Trans-repression is thought to be mainly due to direct interactions between GRα and pro-inflammatory transcription factors such as the activator protein-1 (AP-1) and NF-κB [24]. Trans-activation is explained by the interaction of GRα to specific target DNA sequences, named glucocorticoid-responsive elements (GRE). Among the genes activated by GRα through GRE with anti-inflammatory functions, there are the mitogen activated protein kinase phosphatase-1, the glucocorticoid inducible leucine zipper and tristetraprolin. In addition, the activated GRα can also reduce inflammation at the post-transcriptional (altering mRNA stability), translational (affecting protein synthesis) and post-translational levels (altering protein processing, modification or degradation) (Fig. 2). For example, the expression of cyclooxygenase-2, TNF-α and GM-CSF are regulated by one or more of these postgenomic mechanisms [25].

Increased expression of GRβ has been reported in different inflammatory diseases, including asthma, and nasal polyposis and has been proposed as one of the potential mechanisms explaining GC resistance [26]. The expression of GRβ is higher in nasal polyps than in nasal mucosa epithelial cells and correlates with increased infiltration of inflammatory cells [27]. Although down-regulation of GRα after treatment with glucocorticoids has been reported [28] and could account for secondary steroid resistance, a recent study in patients in patients with nasal polyps has shown that this effect does not occur in vivo [29].

Although GCS play a key role in the treatment of various inflammatory disorders, including chronic upper airway disease, a quite extensive range of potential AE’s is well-described in literature and the chance to develop these effects seems to increase with higher dose and longer duration of treatment [8, 100‐102].

However, few studies have actually addressed the risk of common GCS-induced AE in upper airway disease. Also, most of the studies available on GCS focus on high dose or long-term usage for at least 6 months or even 1 year consecutively, which is mostly less relevant in the upper airway disease patient group.

In the following section, we aimed at summarizing the data of potential short- as well as long-term AE’s of systemic GCS treatments for rhinitis and/or rhinosinusitis in the adult population. Due to the heterogeneity in studies, treatment regimens and patient populations, we classified the side-effects according to the organ-system involved, but no further subdivision was made. When no studies were available for upper airway disease patients, a mention of studies investigating AE’s in similar patients (ophthalmologic, asthmatic) was made. Studies investigating side-effects in children will be discussed separately in the next chapter.

Inflammatory diseases of the nose and paranasal sinuses in children include upper respiratory tract infections, chronic rhinitis, ARS and CRS. ARS is defined as increase of sinonasal symptoms after 5 days of infection or persistent symptoms after 10 days and characterized by the sudden onset of two or more of the symptoms (discoloured nasal discharge, nasal blockage/obstruction/congestion, cough at daytime and night-time) for less than 12 weeks [4]. Bacterial infection is expected when at least 3 symptoms are present among which discoloured discharge, purulent secretion in nasal cavity, severe local pain with a unilateral predominance, fever, elevated C-reactive protein or erythrocyte sedimentation rate, and double sickening (i.e. deterioration after an initial milder phase of illness) [4]. The definition of pediatric CRS differs from adult CRS by the symptom of cough [4] and is defined by the presence of two or more symptoms, one of which should be either nasal obstruction or nasal discharge (anterior or posterior) with/without facial pain/pressure with/without cough, lasting for at least 12 weeks [4]. The diagnosis is confirmed by either nasal endoscopy showing edema, purulent drainage or nasal polyps in the middle meatus or CT scan showing ostiomeatal complex or sinus opacification. Of note, the presence of nasal polyps is much less common in pediatric patients than in adult patients with CRS [126].

Besides clinical consequences, systemic GCS use may also have some health economic implications that should be considered in its benefit-harm trade-off. Generally, the direct costs for systemic GCS are among the lowest quartile of prices of medications available worldwide. However, the indirect costs due to adverse events of (especially long-term, high-dose) systemic GCS use could be more substantial. Two industry-funded studies have assessed the cumulative economic burden of GCS associated adverse events regardless of dose, duration or indication [131, 132]. Manson et al. [131] identified 63 studies in which 21 different GCS adverse events were reported with increased fracture risk, gastric and psychiatric conditions being the most frequent ones. Their economic analysis from the UK perspective revealed that taking oral GCS would result in an additional annual cost of at least £165 for treatment of all steroid related adverse events. One study specifically assessed the economic impact of oral GCS on related fractures where hip, vertebral and forearm fractures costed £10,761, £1976 and £863 respectively. Notably, only three studies focused on patients with allergic rhinitis and/or skin diseases and none specifically on rhinosinusitis. A second review [132] included 47 studies reporting on adverse events of systemic GCS. Subsequently, a cost analysis was undertaken from the US perspective. It was unclear whether any patients with allergic rhinitis or rhinosinusitis were included. Most frequently reported adverse events were psychiatric and gastric conditions, infections and fractures. The authors estimated the potential cost reductions if the daily GCS dose would be reduced. Regarding avoidance of fractures, they estimated that 96 fractures per 10,000 elderly patients could be avoided summing up to $1.76 million ($176 per patient). The findings from both reviews should be interpreted with caution given the heterogeneous and often low-quality and retrospective nature of the studies included and the difficulty in excluding confounding due to underlying disease activity. Besides these two reviews with no particular disease focus, some studies focused on the costs of systemic GCS related adverse events within a specific population such as asthma [133, 134] or rheumatologic diseases [135, 136] and found increased costs in the GCS exposed populations. None were specifically focusing on rhinitis or rhinosinusitis. We conclude that given the limited amount of current evidence, more studies on the economic burden and cost-effectiveness of systemic GCS use in rhinitis and rhinosinusitis treatment are required.

In both rhinitis and rhinosinusitis patients, systemic GCS treatment is in general reserved for those in whom disease control cannot be obtained by baseline medical therapy (intranasal steroids and antihistamine/antileukotrienes for AR [30] and intranasal steroids and antibiotics for ARS/CRS [4]). However, in AR, allergen immunotherapy (AIT) is an alternative option for patients suffering from uncontrolled symptoms. AIT modifies the natural disease course and recent well-performed trials have demonstrated reductions in both symptoms and use of rescue medication in patients with AR for both the subcutaneous as well as sublingual administration route [137]. One study from 1969 compared the efficacy of one depot MP injection with a pre-seasonal administration of an alum-precipitated pyridine extracted grass pollen immunotherapy and found similar results between the two groups in terms of symptom improvement [138]. However, this paper already stated that the potential AE’s of MP do not justify the use of systemic GCS for a condition such as AR. One large Danish registry study including almost 40,000 AR patients actually showed the oral steroid-sparing effect of subcutaneous AIT (SCIT) for seasonal AR with an annual mean of 1.0 steroid injections in patients receiving SCIT versus a mean of 1.6 injections in the non-SCIT group. Of the SCIT-treated individuals, 84% did not need GCS at all after SCIT treatment [139]. Aasbjerg looked at the same registry to compare AE’s and found that AR patient treated with systemic GCS showed more diabetes and osteoporosis than those treated with AIT as mentioned above [107].

For CRS patients, current alternatives for oral GCS during exacerbations consist of antibiotics and when patients remain uncontrolled, sinus surgery is the next step in line [4]. However, studies investigating biological agents that are available for the treatment of asthma and/or other allergic diseases, have shown very beneficial effects in CRSwNP patients [140] but are currently only available for those with severe concomitant asthma.

Gevaert et al. [141] extrapolated results from different studies to compare the efficacy of different treatments in CRSwNP patients. They found a beneficial effect on NP score of doxycycline that was comparable to MP after 8 weeks. Also, omalizumab and mepolizumab treatment had better results on NP score than the oral GCS treatment. Omalizumab and mepolizumab additionally showed better symptom control compared to MP. Currently only data on the oral steroid-sparing effects of mepolizumab and benralizumab in asthma are available [142], but with the increased implementation of these therapies in CRSwNP, studies evaluating the steroid-sparing effect for upper airway exacerbations will be necessary.

When disease control in upper airway disease cannot be obtained with intranasal steroids or other medical treatment prescribed by the respective guidelines, severe cases of AR, ARS, AFRS and CRSwNP can be treated with a short-term course of systemic GCS to improve symptoms. This manuscript provided an overview of the current evidence for the beneficial effects of systemic GCS in the different subtypes of upper airway diseases, as well as in the pediatric age group and aimed at providing recommendations for the specific disease entities.

However, multiple AEs have been widely described and therefore physicians should be aware of the risks associated with oral GCS and make a good risk–benefit assessment prior to prescribing them. In this paper, we summarize these potential AEs; given the current evidence in literature, a clear assessment of the risks associated with oral steroid use in upper airway disease cannot be made. Currently available data show a wide variability in diseases, patients, duration of treatment and follow-up and therefore this topic needs to be addressed in a systematic way in order to provide a substantiated recommendation for the use and dosing of oral GCS in the upper airway disease population.

We can conclude that, although some beneficial effects of systemic GCS have been demonstrated in chronic upper airway diseases such as AR and CRSwNP, systemic GCS should not be considered as a first line of treatment for these disease types.