Introduction
Cerebral abscesses (CAs) are often associated with high morbidity and mortality, and in a large proportion of CA patients, the origin of the infection is unknown [
1,
2]. Identifying the origin is crucial in preventing and treating the infectious disease, as elimination of the primary focus of infection should be a priority to avoid the spread of microorganisms [
3,
4]. From 1982 to 2016, an increasing incidence of CAs has been observed in Denmark [
5], and the number of patients with unknown source of infection has grown [
1]. Consistent reports of oral microorganisms isolated from CA lesions have made odontogenic infections widely recognized as a potential source of infection [
5]. Still, a meta-analysis including studies from 1970 to 2013 identified odontogenic predisposing infections in only 5% of the patients [
3]. However, the included studies showed great heterogeneity precluding further analysis of the predisposing infections described. In recent studies, CAs of odontogenic origin have been reported to be more frequent [
5,
6]. The reasons for this may be that more teeth are preserved nowadays [
6], and there is an increased awareness that oral infections may spread to distant anatomical sites. Because of the increasing incidence of odontogenic CAs, a comprehensive understanding of their pathogenesis has become increasingly important. However, only a few studies have focused on odontogenic CAs. The purpose of the present study is to identify and characterize CAs of odontogenic origin in a well-defined cohort based on clinical, radiographic, and microbiologic findings.
Discussion
In the present study, 57% of the CA patients were characterized as having an odontogenic origin. This proportion is higher than previously reported [
1,
3,
5,
6]. The thorough examination of radiographic material revealed oral pathologic conditions in 11 patients (25%) during hospitalization, in whom predisposing infections were initially not registered. Several other studies have reported a substantial group of patients with CA of unknown origin [
1,
2]; however, only one patient had an unknown origin in this study. This finding suggests that some patients in previous studies categorized with unknown origin may have had an overlooked odontogenic predisposing infection. As oral pathologic conditions do not always exhibit any clinical signs or symptoms [
12], more attention to the clinical dental and radiographic examination during hospitalization is recommended.
The oral pathologic conditions observed in the odontogenic CA group are similar to those reported in previous studies [
11,
13]. They include signs of apical periodontitis and severe periodontitis involving molar teeth with no predilection for mandibular versus maxillary teeth. Also, signs of pericoronitis were present in several patients. Most patients had signs of multiple oral pathologic conditions indicating that neglect of oral care could be a potential issue in some of these patients. The identified oral pathologic conditions were all inflammatory diseases that are expected to increase the risk and load of bacteremia, leading to the dissemination of oral microorganisms causing systemic and distant site infections [
14‐
16]. Dental procedures performed before admission were not systematically registered in the medical journals and were therefore not included for further analysis in this study. However, a recent study found no increased risk of developing CA after invasive dental procedures [
17]. Eleven patients characterized as CA with odontogenic origin had solely oral pathologic conditions revealed on PRs, and these were not examined during hospitalization by an oral and maxillofacial surgeon. Although PRs compared with other radiographic techniques have a limited value detecting some oral pathologic conditions [
12,
18], a high agreement among the examiners in this study validated the actual presence of oral pathologic conditions on admission in these 11 patients. In the non-odontogenic group, 8 patients did not have radiographic material of the mandibular and maxillary region, as an obvious infectious focus was found at an early stage during hospitalization. In both groups, four patients did not have an independent radiographic examination, as CT-material were evaluated in cooperation between the examiners. The evaluation of CT scans was done without the examiners having access to background information on the patients; however, it is a limit to the radiographic evaluation.
Sinusitis was the second most common predisposing infection, and attention should be paid to the fact that maxillary sinusitis may be odontogenic as well. Odontogenic sinusitis is a prevalent and underdiagnosed disease caused by apical periodontitis or iatrogenic extrusion of foreign bodies into the maxillary sinuses and is strongly associated with unilateral maxillary sinusitis [
19,
20]. Therefore, CA patients with unilateral maxillary sinusitis should have an oral examination conducted as odontogenic sinusitis usually requires combined dental and medical treatment [
21].
In four patients characterized as having odontogenic CA, predisposing risk factors were present. In three out of four patients, the predisposing risk factors were related to prior surgical interventions or head trauma. However, the pathogens isolated were microorganisms predominantly found in the oral cavity and the risk factors were not attributed to CA formation in the medical records. One of the patients received sinus surgery prior to hospitalization due to oral pathologic conditions, and one patient had a patent foramen ovale, which is not per se the reason for CA formation. Therefore, all four patients were evaluated to be justifiable and characterized as odontogenic CA.
Diabetes has previously been suggested as a risk factor that may exacerbate odontogenic infections [
22], and T2D with poor glycemic control is well known to have a bidirectional relationship with periodontal disease, due to negative effects on the immune system [
23]. Based on the findings from the present study, diabetic patients should pay extra attention to preserve healthy oral conditions, as T2D patients were overrepresented among patients with odontogenic CA (
p = 0.014). Also, in 5/6 patients, the HbA1c level was above 7.7 mmol/L indicating that poor glycemic control may be correlated with odontogenic CA formation.
Microorganisms from the SAG were the most prevalent pathogens isolated in the present cohort. The SAG consists of
Streptococci constellatus,
Streptococci intermedius, and
Streptococci anginosus, and the group is frequently commensals of the oral cavity, but also isolated from the upper respiratory, gastrointestinal, and genitourinary tracts [
24]. These bacteria are well-recognized for causing central nervous system infections, and in recent decades, the incidence of CAs with SAG bacteria has increased [
25]. They are often isolated from odontogenic infectious lesions including periapical abscesses, periodontal abscesses, and pericoronitis [
26‐
30]. Also,
S. constellatus and
S. intermedius found in dental plaque have been associated with the development of periodontal disease [
31]. A recent study from Sweden concluded that
S. anginosus alone, as a single species, accounted for over 20% of oral bacterial infections [
30], and a previous study showed a very close similarity between brain and oral isolates of SAG in a CA patient with periodontitis, where the use of highly discriminative phenotypic and genetic fingerprint techniques could not exclude that microorganisms involved in CA formation came from the periodontal pocket [
32]. However, it is important to bear in mind that microorganisms from the SAG, especially
S. anginosus, are commonly isolated from gastrointestinal and genitourinary tract infections as well [
24]. Though, none of the patients in the odontogenic group had any signs of infections in addition to the oral cavity.
The SAG bacteria use several strategies to invade and colonize host cells and tissues. However, the understanding of their virulence factors and especially their regulation is limited [
33]. In biofilms, the regulation of virulence genes is controlled through quorum sensing [
34], and it has been suggested that synergy between SAG bacteria and oral anaerobes results in enhanced growth of SAG bacteria [
28].
The analyses of the relationship between the suspected reason of CA development and the microbiological composition of the brain pus indicate that when bacteria from the SAG or other oral bacteria are isolated in the brain pus, a thorough dental examination should be conducted. This corresponds well with previous findings, where isolation of SAG from brain pus has led to the classification of the CA as odontogenic origin [
35,
36]. While several odontogenic CA patients in the present cohort had a monomicrobial infection of SAG isolates or other oral microorganisms (Supplementary
S3), other studies suggest that odontogenic CAs are predominantly associated with polymicrobial infections [
2]. However, as the oral cavity harboring > 700 bacterial species of which only 68% have been cultured [
24], culture-based methods cannot be expected to detect all slow-growing and fastidious oral microorganisms. Additionally, culturing may be impaired by the administration of antibiotics preoperatively and by the nutritional media used for sample transportation and general conditions of transportation [
37]. In polymicrobial samples, conventional 16 s rDNA sanger sequencing is also inadequate to discriminate among all taxa, and some samples may produce false-negative results due to low bacteria concentration [
37]. Recent studies using metagenomic and next-generation 16 s rDNA sequencing methods disclosed a wide variety of oral microorganisms previously unknown to be involved in CA development, and a larger polymicrobial proportion than previously thought was found [
35‐
37]. Therefore, a monomicrobial isolate should be interpreted with caution and should not rule out a potential odontogenic origin.
In the non-odontogenic group, 6/11 patients had radiographic pathologic conditions. This finding is not surprising, as there is a high prevalence of asymptomatic oral infections in the general population [
38,
39]. It is presently unknown why a limited proportion of these conditions develop into CA. Future investigations should focus on the host-pathogen interaction, as an aberrant immune response to bacterial infections in some patients may allow CA development. The identification of any immune deficiency in combination with specific odontogenic infections may assist in developing preventive measures to reduce the risk of developing CA.
This study has several limitations. The main limitation is the risk of ascertainment bias as there could be an unavoidable risk of favoring the characterization of patients with odontogenic origin. However, the risk was kept at a minimum due to the evaluation of intra- and inter-rater reliability during the radiographic assessment and the use of pre-defined criteria when characterizing CA patients with odontogenic origin. Even though CA patients were documented in the DASGIB database, some patients were referred from other hospitals whereby clinical information from hospital admission in certain cases were limited. Due to the limited focus on oral health during the hospitalization, the main recordings of these parameters were based on radiographic information, on registrations from the oral maxillofacial surgeon that evaluated a selected proportion of the patients, and on occasional remarks in the medical records. The cohort examined is well-defined, and the strength of this study includes the continuous registration of data in a predefined registration form in the DASGIB database.
However, when using the three criteria to characterize odontogenic CA patients, there is a potential risk to overestimate the possible frequency of CA with odontogenic origin, as microorganisms normally present in the oral cavity might be involved in infections external to the oral cavity.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.