Unlike Periodontitis, Caries Does Not Associate With Intracranial Aneurysms or Aneurysmal Subarachnoid Hemorrhage


 Background:Periodontal infections have been associated to the formation and rupture of intracranial aneurysms (IA). In this study we investigated whether also caries (tooth decay) associates to IA pathology.Methods:A total of 166 patients with either IA or aneurysmal subarachnoid hemorrhage (aSAH) underwent oral examination in Kuopio University Hospital and Tampere University Hospital. Findings were compared to age, gender and geographically matched controls acquired from cross-sectional Health 2000 Survey. This study consisted three sequential steps. First step was a comparison of the caries prevalence and number of missing teeth in IA and aSAH patients with the control population, second step was a multivariate analysis with demographic factors and third step was a prospective 13-year follow-up of participants with any caries or missing teeth.Results:In logistic regression adjusted for known risk factors and demographic data, caries (OR: 0.412 95%Cl 0.2-0.9, p=0.028) was associated with lack of IAs, while age (OR:1.029 95%Cl 1.0-1.1 p=0.039), current smoking (OR:2.7 95%Cl 1.4-5.1, p=0.003) and severe periodontitis (OR:4.3 95%Cl 2.3-12.5, p<0.001) associated with IA formation. Severe periodontitis associated also to aSAH in logistic regression (OR: 5.4, 95%Cl 1.9-15.5, p=0.002). In the cox-regression, severe periodontitis at baseline increased the risk of aSAH (HR: 11.9, 95%Cl 1.2-114.7, p=0.032) during a 13-year follow-up.Conclusion:Unlike periodontitis, caries does not increase the risk of IAs and aSAHs. However, cariogenic bacteria may participate to IA pathology by disseminating to circulation via inflamed gingival tissue.


Introduction
Unruptured intracranial aneurysms (UIA) are frequent: They are found in approx. 3% of the past middle age population as mostly asymptomatic cerebral artery lesions. [24]. Despite the rather high prevalence of UIAs, intracranial hemorrhage resulting from IA rupture is a relatively rare event (incidence approx. 10/100 000 in most countries) [18]. This form of intracranial hemorrhage known as aneurysmal subarachnoid hemorrhage (aSAH) has, however, a particularly sinister prognosis with a mortality reaching almost 50% and with many of the survivors left signi cantly disabled [23]. The best treatment for aSAH is to prevent it, a goal for which it is paramount to predict which UIA is going to eventually rupture and who will develop UIAs in the rst place.
Chronic in ammation or in ammatory cell mediated remodeling of the cerebral artery has been shown to be a crucial mediator of IA formation, as well as of the IA wall degeneration that eventually leads to rupture [3,6,7,12]. The presence of bacterial derived DNA in the IA walls, as well as the expression of Tolllike receptors that react to bacterial components and activate the immune system has been shown earlier [21]. This strongly implies that dental pathogens play a role in the in ammation mediated artery and aneurysm wall remodeling that leads to IA formation and aSAH. In addition, a prior study using RT-qPCR demonstrated that this bacterial DNA originates from several dental pathogens [21].
Steptococcus mutans is a classical dental pathogen involved in caries, a highly prevalent dental infection that leads to formation of cavities in teeth, and if untreated, to in ammation and necrosis of the tooths nerve system, pulp. Eventually these conditions may lead to root canal treatment or extraction of the affected tooth. Certain S. mutans strains like the Cnm and Cbm proteins possess proteins capable of binding collagen and are linked to aggressive form of caries [15]. S. mutans expressing collagen-binding protein (CBP) has been previously shown to be more prevalent in oral samples of cerebral haemorrhage patients [17] and the persons affected by aSAH or other forms of intracranial hemorrhage [11]. CBP of S. mutans has been shown in animal models to predispose to intracranial hemorrhage through affecting the cerebral artery wall directly and can be detected in cerebral haemorrhage tissue after oral administration [17]. It seems possible that this pathogen, so important in caries, would also play role in IA formation and rupture.
Since it was recently observed that gingivitis and periodontitis, i.e. chronic infection of the gums and tooth supporting tissues, associates with the risk of developing IAs and eventually aSAH [9,22] and prior literature suggest involvement of also the caries pathogen S. mutans in IA formation and rupture, we now investigated whether caries is associated with the IA formation and the eventual risk of aSAH from IA rupture.

Materials And Methods
This study was performed in three sequential steps. First step was a comparison of the caries prevalence and number of missing teeth between IA and aSAH patients and the control group. Second step was a multivariate analysis between demographic factors, caries, missing teeth and periodontitis and IA and aSAH prevalence and third step was a prospective 13  . Patient selection was random due to the limited availability of an examining dentists (JH and MP) and interexaminer validation was not done. In KUH IA patients caries lesions were common but caries diagnostics could not be done due to the lack of panoramic x-rays and KUH IA patients were therefore not included in the analysis. In TaUH IA patients, number of teeth and caries was examined from panoramic x-ray. Caries was diagnosed when the caries lesion was clearly extending to dentin (a bone-like matrix under tooth enamel). Periodontitis was diagnosed taking into account the deepest periodontal probing depth, categorized as follows: < 4 mm no periodontitis, 4-5 mm periodontitis, and = 6 mm severe periodontitis [9,22]. For the KUH and TaUH IA patients, clinical data including known risk factors (age, gender and current smoking) was collected from the medical reports and a personal interview.
Case-control comparison of IA patients A previously published prospectively collected cohort of 8028 adult Finns of whom a total of 5144 participants underwent a baseline oral examination. From this group of 5144 participants, we selected the geographically matched controls from TaUH area (n=340) as our control group for the studied IA and aSAH patients of Tampere University Hospital. Demographics and risk factors for IA were collected using a questionnaire as described previously [2]. To identify possible prior uIA or aSAH before baseline the national registry for hospital discharge diagnosis (HILMO) was searched for ICD-10 codes I67.1 or I60.0-I60.9. Also procedure codes were searched for surgical or endovascular procedures according to the Nordic Classi cation for Surgical Procedures.

Prospective follow-up of Health 2000 -survey participants
Follow-up data was collected from HILMO registry after the baseline examination but before December 31 st 2013 as previously described [9]. Cox-regression was used to calculate the hazard ratio for IA formation and for aSAH during the 13-year follow-up after the baseline examination.

Statistical analyses
Data was presented as frequencies or medians with ranges. Fisher's exact test or chi-square tests were used to compare categorical variables and Mann-Whitney U test for continuous variables. Multivariate logistic regression and Cox-regression were performed as described above. Results from both regression analyses were presented as odds-or hazard ratios with 95% con dence intervals. SPSS 22.0 statistical software (IBM) was used and a p value < 0.05 was considered as signi cant.

Results
Missing teeth and prevalence of caries among UIA and aSAH patients UIA patients from both KUH and TaUH cohorts had a median of 2 missing teeth and aSAH patients had median of 1 missing teeth. After dividing the study group to quartiles according to the number of missing teeth, there was a trend towards higher number of missing teeth among UIA and aSAH patients, but this remained non-signi cant. (data not shown).
In the subgroup of TaUH patients, at least one dentin caries lesion was found in 18.3% of the UIA patients (Table 1) and 30.0% of the aSAH patients (Table 2). 25.6% of the controls from the same region had at least one dentin caries lesion (Tables 1 & 2 Table 3, models 1&2) also in the TaUH cohort. On the contrary, number of missing teeth was not associated with uIAs nor aSAH (Table 3 Model   1&2).
Unlike severe periodontitis, caries did not associate with later aSAH in a 13-year follow-up.
In a Cox-regression model adjusted for age, gender, current smoking, caries, peridodontitis and missing teeth at baseline, only severe periodontitis at baseline increased the risk of aSAH (Table 4). Patients were identi ed by having both ICD10 and procedure code relevant to aSAH.

Discussion
In this study caries associates inversely with the formation of IAs, but not with aSAH. Caries is the most common disease globally affecting nearly 3.5 billion people worldwide [13]. Despite the fact that caries is more prevalent in high-socioeconomic states [13], our study of high-socioeconomic population, did not associate caries to a greater risk of IA and aSAH. There are several explanations for the lack of association: rst the nature of caries is a local infectious disease of the tooth structure with systemic responses only when the caries reaches the tooth pulp. Secondly, caries is most often treated with llings before severe conditions, such as in ammation of tooth pulp or periapical abscesses occur. However these caries-related conditions remain a challenge to oral health care providers. Thirdly, caries does not predispose to bacteremia contrary to periodontal diseases of the gingival tissues.
The association studies of oral bacteria with vascular diseases, including sIA/aSAH concentrate mostly on periodontal bacteria. Regarding cariogenic bacteria, a Japanese study [11] linked speci c cnm strain of Streptococcus mutans to sIA. This speci c strain promotes platelet aggregation inhibition and matrix metalloproteinase-9 activation which can logically be linked with sIA formation and rupturing. Although in our study caries did not associate with sIA disease, the nding of Inenaga et al may be explained by the poor oral hygiene predisposing to a poorer periodontal status. Via in amed periodontal tissues this speci c S. mutans strain could disseminate to circulation and to IA walls. This explanation is, however, highly speculative.

Other dental infections as risk factors for IA formation and subsequent aSAH
Whereas caries does not appear to associate with IA and aSAH, gingivitis and periodontitis, however, have been shown to associate to both IAs and aSAH [8- 10,22]. Similar association of periodontitis to IAs and aSAHs that we have previously reported in the KUH cohort [9], was also clear in this study with TaUH patient cohort with geographically matched healthy controls. Furthermore, previously a plethora of studies have associated periodontitis to cardiovascular diseases [14,20] and this further demonstrates the crucial role of gingival pocket health as the border between systemic circulation and oral cavity. The association between oral infections and vascular diseases could be explained by either (i) direct effect on local bacterial infection on systemic low grade in ammation response, (ii) bacteremia and subsequent distant local infections (such as endocarditis) or (iii) similar genetic background of the diseases.
Periodontal pathogens correlate to diseases/medical states of which pathogenesis is strongly linked to collagen degradation caused by MMP activation in different sites of the human body [4,11,25], such as pre-term birth and low birthweight [5], cardiovascular diseases [16] and periodontitis [19]. A single speci c nding that further explains the association was introduced by Aarabi and colleaques in 2017. They described common genetic background of periodontal destructive and vascular destructive states [1]. The involvement of shared genetic background of the diseases could also explain the associations: Some people are prone to both periodontal disease and cardiovascular diseases, and the bacteria invade the vascular wall if they are given the chance by genetic factors. Pathomechanisms of caries are not similar with these collagen degradative states seen in periodontal diseases.
Since periodontal diseases associate with sIA disease and increase the risk of aSAH in the long term, but other oral diseases, to our knowledge, do not, the effect and association might be explained by the nature of the disease itself rather than periodontal disease being only an "innocent co-variant" re ecting for example poor socioeconomic status to which many IA risk factors associate to. A prior study on the bacterial DNA present in IA wall tissue have not been able to discriminate whether periodontal or cariogenic pathogens associate with IAs, but rather DNA of pathogens related to both groups are found in the IA wall [21]. This discrepancy between the bacterial DNA ndings and the clinical nding that nevertheless only periodontal disease associates with IA formation or rupture could be explained by the bleeding wound surface in in amed gingival pocket forming a gateway for oral bacteria to enter the systemic circulation, and eventually the IA wall. Further studies on this hypothesis are warranted.

Limitations of the study
Our study has some limitations due to its design. First, we could not recruit all IA and aSAH patients to the study due to patients denial to participate or due to tight schedule of the examining dentists. The use of healthcare registry data has always a source of bias. Also, our comparison between KUH and TaUH patients to participants of Health 2000 -survey has its limits, since there is a 15-16 year gap between oral examinations. Also, single person performing the oral examination (JH in KUH and MP in TaUH) could lead to over-or underestimations of oral conditions.

Conclusion
Unlike periodontitis, caries does not associate with the IA formation or aSAH. However, cariogenic bacteria may participate to IA pathogenesis, but further research is needed Declarations Funding: This study was funded by a research grant from the Finnish Medical Foundation (Dr. Frösen) and by a research grant Finnish Dental Society Apollonia (Dr. Hallikainen).
Con icts of interest: Authors declare no con ict of interest to this study.
Availability of data and material: Due to the nature of this research and GDPR policy, participants of this study did not agree for their data to be shared publicly, so supporting data is not available.