Acute mastoiditis and intracranial complications in children

Overview

Objectives: Characterisation of clinical manifestations in children who had acute mastoiditis (AM) -related intracranial complications (ICCs) and to determine the incidence of ICCs in the study group. To define children with a higher risk of ICCs and gain new information on aetiological microbial agents of AM. Methods: A retrospective analysis of 137 paediatric patients with AM treated at a tertiary centre using standard dia­gnostic and therapeutic protocol between 2002 and 2019. Results: Altogether 137 patients with AM were hospitalised at our centre between 2002 and 2019. During this time, the occurrence of ICC in children with AM was low (n = 3, incidence 2.19%). Due to the low number of patients with ICC with AM in our patient group, we were unable to define a specific group of patients with a higher risk of ICC development in AM. Despite this fact, the average value and median of CRP were considerably higher in patients with AM-associated ICCs. The most frequent aetiological agent was Streptococcus pneumoniae. Conclusions: The low incidence of ICC in patients with AM in our patient group might be caused by routinely performed paracentesis in developed acute otitis, rational antibio­tic therapy and the good availability of an ENT specialist in the Czech Republic. The most common aetiological agent of AM was Streptococcus pneumoniae.

Keywords:

Streptococcus pneumoniae – acute mastoiditis – intracranial complications – aetiological agent – antibio­tic therapy – antromastoidectomy

Introduction

Acute mastoiditis (AM) is the most com­mon inflammatory complication of acute otitis media (AOM) in children. The incidence of AM drops significantly after the introduction of antibio­tic therapy [1]; the incidence of intracranial complications (ICCs) associated with AM ranges from 5 to 29%. These ICCs can include perisinus empyema, epidural abscess, sigmoid sinus thrombosis and thrombophlebitis, subdural empyema, brain and cerebellar abscess, meningitis or otitic hydrocephalus [2–6]. Such complications may present at the time of dia­gnosis of AM, however they are also capable of developing later despite adequate therapy. Considering the potential for serious consequences of ICCs, early dia­gnosis is crucial since delayed dia­gnosis increases patient morbidity and mortality [7, 8]. Imaging methods are an important part of the dia­gnosis of ICCs in AM. MRI is the most suitable imaging method, however this examination takes more time than CT and in children must be performed under general anaesthesia, therefore it is not often indicated [9]. CT has become the gold standard in the dia­gnosis of ICCs, however radiation dose on the brain remains pertinent to discussion. Therefore, in many centres CT is indicated individually – based on clinical symptoms and laboratory tests. Currently there are no standardised guidelines and indication is thereby in the hands of the attending physician. Some studies have attempted to further elucidate a clear combination of clinical parameters that might distinguish children with a higher risk of ICCs [5, 10]. Antibio­tics are the key factor in treatment of AM with ICCs. It must also be noted that nowadays, some changes have been observed in the microbial aetiological agents of AM [2].

Aims

To characterise the clinical manifestation in children who had intracranial complications (ICCs) related to AM and determine the incidence of ICCs in our patient group. To define children with a higher risk of ICCs and gain new information on aetiological microbial agents of AM.

Methods

Analysis of all paediatric patients with AM examined at our centre between 2002 and 2019 using standard dia­gnostic and therapeutic protocol. None of the patients was treated for chronic otitis or had a cochlear implant.

The dia­gnosis of AM was based on clinical symptoms: otoscopic findings on the tympanic membrane with signs of acute otitis (bulging of the tympanic membrane, tympanic membrane after spontaneous perforation with otorrhoea or tympanic membrane after paracentesis), inflammatory changes in retroauricular area (tenderness, redness, swelling and/or subperiosteal abscess) and protrusion of the auricle. All the patients had the above-mentioned symptoms at the time of the initial examination. The material for microbial examination was acquired from ear secretion (after para­centesis or spontaneous perforation), during surgery (antromastoidectomy) or after peroperatively performed myringotomy. The patients were divided into two groups, with and without ICCs. The aim was to identify the characteristics of the two groups in order to define the group with a higher risk of ICCs.

The Fischer exact test at a level of significance p = 0.05 was used for statistical analysis. TIBCO Software Inc. (2018) was used for statistical analysis. Statistica (data analysis software system), version 13, http: //tibco.com.

Results

During 2002–2019, 137 paediatric patients were treated with AM, comprising 80 boys (58%) and 57 girls (42%). The right ear was affected 76 times and the left ear 62 times. In one case both ears were affected. The age of the patients ranged from 2 months to 17 years, where the average age was 4 years.

All the patients had clinical symptoms of AM. The patients were divided into two groups according to the presence of ICCs in AM. The characteristics for both groups (AM with ICCs and without ICCs) are listed in the table (Tab. 1). Due to the low number of patients with ICCs in the observed patient group, no valid analyses in terms of group differences can be made.

The average value of CRP at the time of admission was 81.09 mg/l and leucocytosis of 16.51 × 10⁹/l (24 patients did not have values of leucocytosis and 13 did not have CRP values). In patients with AM with ICCs, the initial value of CRP was significantly higher than in patients without ICCs. Conversely, the leucocytosis value was notably raised in patients without ICCs (Tab. 1). Due to the low numbers of patients with AM with ICCs in our patient group, the inflammatory parameter values cannot be statistically evaluated satisfactorily (low test strength).

The result of the microbio­logical examination was available in 95 (69.34%) patients (96 cultivations altogether, one patient with bilateral mastoiditis). In case two or more bacterial pathogens were cultivated, only the causative pathogen was considered. A positive cultivation result was found in 72 (75.79%) patients, negative in 23 (24.21%) patients. Overall, 14 out of 23 (60.87%) patients with negative cultivation results had been treated with antibio­tics before AM developed, while 24 out of 72 (33.33%) patients with positive cultivation had been treated by antibio­tics before AM developed (Tab. 2). Fifty-nine patients were treated by antibio­tics before admission to hospital. Information on the type of antibio­tics was available in 55 out of 59 cases. The most commonly used were aminopenicillins, in 17 out of 55 patients (31%), then cephalospo­rins of the II^nd generation in 13 out of 55 cases (24%) and macrolides in 12 out of 55 cases (22%).

In the entire patient group, the most common pathogen identified was Streptococcus pneumoniae (25 cases out of 72 positive cultivations). Streptococcus pneumoniae was also the most commonly cultivated pathogen in the group of patients not treated by antibio­tics before admission to hospital (19 out of 48 cases), the second most common was Grp. A Streptococcus pyogenes (12 out of 48 cases). In the group of patients who were treated by antibio­tics before admission to hospital, the most common aetiological agent was Streptococcus pneumoniae (6 cases out of 24) and Pseudomonas aeruginosa (6 cases out of 24). The patients treated with antibio­tics before admission to hospital had a statistically lower number of positive cultivations. Conversely, in the patients treated with antibio­tics before admission to hospital there was a statistically different prevalence of other than observed (in our other bacteria article group) pathogens (Tab. 1, 3, Fig. 1).

Antibio­tic therapy was initiated in all hospitalised patients. In 134 cases out of 137, antibio­tic therapy was administered intravenously, in 3 patients it was administered per os. Antibio­tic monotherapy was indicated in 124 patients, in some cases (13 patients) in combination with another antibio­tic (the most common combination was clindamycin with gentamicin). In monotherapy, lincosamides were selected the most frequently, followed by II^nd generation cephalosporines and protected (combined with b-lactamase inhibitor) aminopenicillins.

Seventy-nine patients required surgical therapy. Transcortical antromastoidectomy (by retroauricular approach with myringotomy) was performed in 74 cases. Revision antromastoidectomy was performed in 5 cases. Subperiosteal abscess was the most frequent complication of AM. It occurred in 43 patients. Intracranial complications occurred in three patients with AM and subperiosteal abscess. Peri­sinusoidal abscess was the most frequent intracranial complication, accompanied by incomplete thrombosis of the sinus sigmoideus and transversus, extending to the jugular vein in one case. An epidural abscess occurred in one patient.

All patients with ICCs associated with AM underwent immediate intravenous antibio­tic therapy and surgical treatment of the temporal bone inflammation (transcortical antromastoidectomy with myringotomy). Anticoagulant therapy was indicated in case of incomplete thrombosis of the sinus sigmoideus and transversus, extending to the jugular vein.

Discussion

Our data show a low incidence of ICCs associated with AM during the monitored period. The incidence of ICCs associated with AM in our group of patients (2.19%) is lower in comparison with the previous data (5–29%). The confidence interval found in our study was 0.3–17.6%, compared to the incidence data from literature (5–29%) [5].

The treatment algorithms of acute purulent otitis media, which usually precedes acute mastoiditis and complications thereof, differ in various countries. The algorithms differ in the frequency of antibio­tic use and indication for para­centesis. Paracentesis remains a standard treatment of the advanced stage of purulent otitis media in the Czech Republic, in contrast to other countries where this approach is gradually being abandoned. One reason is lower patient burden as well as the fact that in some countries the dia­gnosis and treatment of acute otitis media are performed by paediatricians [11]. The lower frequency of ICCs may be attributable to the abundant availability of ENT care in the Czech Republic. Currently there is 1 ENT specialist per 10,000 inhabitants (our calculation based on data of the Czech Society of Otorhinolaryngology and Head and Neck Surgery from 2020).

Nowadays, conservative surgical therapy of AM with subperiosteal abscess (retroauricular incision and myringotomy or ventilation tube insertion) is preferred as an alternative to antromastoidectomy, if ICCs and suspicion of Fusobacterium necrophorum are not present [12, 13]. At our department, antromastoidectomy is considered a routine part of the therapeutic guidelines of AM, together with intravenous application of antibio­tics and tympanic membrane paracentesis, taking into consideration the risk of extra and intracranial complications of acute otitis media with mastoiditis [11].

Patients with ICCs without clinical signs of acute mastoiditis (inflammatory changes in the retroauricular area and protrusion of the auricle) were not included in the study.

A dia­gnosis of ICCs associated with AM can be easy in typical and developed symptomatology, however sometimes the symptoms can be subtle or altered by antibio­tic therapy and therefore the dia­gnosis can be difficult. Imaging methods are an important step in the dia­gnostic procedure [4, 10]. At our department, as is the case at many foreign centres, indication of CT or MRI is considered individually – based on clinical signs and laboratory tests. The low number of patients with AM-associated ICCs during the observed period does not enable valid statistical analyses of clinical parameters and laboratory test results in order to define patient groups with a higher risk of ICCs. In a similar study published in 2019 (Mansour et al [5]) (Tab. 4), the authors state that there is a 22.5 times higher risk of development of AM-associated ICCs at the time of admission if the patient had CRP over 93.5mg/l and was treated by antibio­tics before AM developed. In this group of patients CT is recommended to exclude ICCs [5]. In our patient group with AM-associated ICCs, the CRP level ranged between 93.3 and 217.7 mg/l. The average value and median of CRP were considerably higher in patients with AM-associated ICCs (Tab. 1).

The type and virulence of microbes is an important factor in infection transmission. In our patient group, the most frequently identified pathogen was Strep­tococcus pneumoniae (31.58% of all cultivated pathogens), followed by Streptococcus pyogenes gr. A (17.11% of all cultivated pathogens). There was no statistically significant difference in pathogen distribution between the group of patients treated by antibio­tics before hospitalisation and the group of patients who were not treated by antibio­tics before hospitalisation.

In the group of patients who were not treated by antibio­tics before hospitalisation, the pathogen was identified in 66.67%. Patients treated by antibio­tics before admission had a statistically lower number of positive cultivations. In this group, the pathogens were identified in 33.33% of the cases; furthermore there was also a statistically higher occurrence of other than observed (typical) pathogens (other bacteria group). The other bacteria group includes sporadically occurring bacteria which were cultivated (alpha-haemolytic streptococcus, aerobic spore-forming microbes, anaerobic G+ rods, S. Aureus, Staph. coagulase negative and others) and can be a part of the common flora of the auditory canal.

Our study, as well as other corroborative studies, identified Streptococcus pneumoniae as the most frequent aetiological agent of AM [3, 4, 12]. Foreign studies discuss the influence of the pneumococcal vaccine on the distribution of AM pathogens. In the Czech Republic, vaccination against Streptococcus pneumoniae was introduced in 2010 and is optional. During the first few years, the vaccination coverage in children aged less than 1 year (at least 1 dose) was around 80%; in the following years it had a decreasing tendency and in 2017 it reached (in the same age category) almost 68% (according to data available from the General Health Insurance of the Czech Republic) [14].

Some aggressive and antibio­tic- -resistant bacteria strains are often responsible for intracranial complications. Several recent papers have shown increased incidence of Fusobacterium necrophorum infection related to a more aggressive course of AM that was frequently causing ICCs in children [5, 15, 16]. We observed a single case of AM caused by Fusobacterium necrophorum connected to ICC in our dataset. We have not observed increased incidence of the pathogen in our study. Additionally, we did not find any resistant strains in our group.

Conclusion

We believe that the low incidence of ICCs might have been caused by routine indication of paracentesis in advanced acute otitis, rational antibio­tic treatment and the good availability of ENT care in the Czech Republic.

Due to the low number of patients with ICC with AM in our patient group, we were unable to define a specific group of patients with a higher risk of ICC development in AM. In spite of this fact, the average value and median of CRP were considerably higher in patients with AM-associated ICCs.

Streptococcus pneumoniae was the most frequent pathogen related to AM in our group of patients. A higher incidence of complications, especially ICCs, must be taken into account in cases where Fusobacterium necrophorum infection is suspected or confirmed.

Acknowledgements

Thanks and gratitude to MUDr. Mgr. Bc. Zuzana Balážová and Sumeet Gulati (student of Masaryk University of Medicine in Brno) for their English language translation and correction.

Conflict of interest

The authors declare that they have no conflicts of interest.

Funding

Supported by project MUNI/A/1118/2020 (Masaryk University Brno, Czechia).

ORCID authors

A. Kaliariková ORCID 0000-0003-2175-4787,

K. Perceová ORCID 0000-0002-3043-4112,

J. Machač ORCID 0000-0002-7951-5650,

M. Jurajda ORCID 0000-0003-1017-4357,

M. Urík ORCID 0000-0002-2872-185X.

Submitted: 2. 2. 2021

Accepted: 5. 3. 2021

Milan Urík, MD, PhD.

Department of Paediatric Otorhinolaryngology, University Hospital Brno and Faculty of Medicine, Masaryk University Brno

Černopolní 9

613 00 Brno, Czech Republic

urik.milan@fnbrno.cz