The impact of a needs-oriented dental prophylaxis program on bacteremia after toothbrushing and systemic inflammation in children, adolescents, and young adults with chronic kidney disease

Despite extensive progress in the treatment of patients affected by chronic kidney disease (CKD), morbidity and mortality still remain unacceptably high [1]. One of the major causes of death is cardiovascular diseases (CVD) [2, 3].

Highest mortality rates in patients on chronic dialysis have been linked to increased inflammatory blood markers for more than 20 years now [4]. In past research concerning the most reliable biomarkers to predict cardiovascular outcome and mortality in CKD patients, the best evidence has been found for C-reactive protein (CRP) and interleukin-6 (IL-6) [5]. There is evidence that elevated CRP levels ≥ 5 mg/l are strongly associated with calcification of coronary arteries and aorta, which are already present in up to 92% of young adults with childhood-onset chronic kidney failure [6].

Among other factors, poor oral health in patients suffering from CKD is one important, yet often underestimated, source of chronic inflammation [7, 8]. Examining the oral condition in patients with CKD, studies demonstrate there is a much higher prevalence of gingivitis, plaque accumulation, attachment loss, enamel hypoplasia, and gingival bleeding or gingival hyperplasia compared to systemically healthy controls whereas caries prevalence remains low, even when kidney function decreases [9, 10].

There is strong evidence that oral bacteria pass to the patients’ blood not only during invasive dental procedures such as tooth extractions, but also due to daily routine activities such as toothbrushing or flossing, with a bacteremia risk of up to 78% [11]. Children requiring conservative dental restorations show 38.5% bacteraemia after toothbrushing. Viridans streptococci, part of the oral cavity and playing a role in the development of infective endocarditis, could be detected in more than 50% of the samples [12]. Within 15 min after dental intervention, the immune system of healthy patients eliminates the bacteraemia [13].

Bacteria can enter the bloodstream through different pathways. They can enter via periapical lesions via the root canal or through the transitional epithelium passing the gingival tissue. Therefore, it seems plausible that poor oral health (e.g., gingivitis) and trauma during dental intervention correlates with the risk for bacterial transfer [14].

In addition to increasing the risk for bacterial translocation, poor oral health also contributes to the development and preservation of chronic systemic inflammation. In their meta-analysis, Paraskevas et al. stated that patients with periodontitis showed an increase in CRP by an average of 1.6 mg/l [14]. There are further studies demonstrating the correlation between poor oral health and increased intima media thickness as an example of the link between oral and cardiovascular pathology [15, 16].

Taking into account all these factors, the development of interventional studies to examine the effect of restoring oral health on systemic inflammation is necessary. Some argue that non-surgical dental intervention can restore oral health but has no significant impact on systemic inflammation [17, 18] while others observe a decrease in high sensitivity CRP (hsCRP) [19, 20]. Tonetti et al. demonstrated that intensive periodontal care initially results in an increase of proinflammatory biomarkers such as CRP and IL-6 combined with a decline in endothelial function, but that there is a significant positive impact on these factors 60 days after treatment. They state that there is a positive correlation between the degree of improvement in oral health with the improvement of inflammatory condition and endothelial function, so dental care becomes a promising approach in improving long-term morbidity and mortality [21].

We hypothesized that an adequate demand-based dental prophylaxis could not only restore poor oral health in patients with CKD but also lower the prevalence of transient bacteremia after toothbrushing and proinflammatory biomarkers seen as parameters of systemic microinflammation. Designing a pilot study, the main focus was on feasibility of our approach in a real-world cohort of patients with CKD.

The present pilot study was not able to detect a statistically significant change in the prevalence of transient bacteremia by using BACTEC^TM blood culture bottles after toothbrushing in patients with CKD following intensive dental prophylactic treatment. In addition, there is no proof for a positive impact on CRP levels as a marker for chronic inflammation. The gingival index (GI) in the IP group decreased by 90% (GI 0.09; SD 0.09; p = 0.001), resulting in an almost healthy gingiva in this group after intensive prophylactic measures. Gingivitis in the TAU group improved by only 15% (GI 0.97; SD 0.58), proving superiority of IP concerning oral health.

Our study resulted in a detection of bacteria after toothbrushing in 9.5% of the analyzed samples, affecting 26% of the participants throughout the study period. In addition, two of the three participants who had to be excluded from analysis due to antibiotic prophylaxis showed transient bacteremia throughout the study period, as well. Ensuring this inclusion criterion was rather difficult in our specific cohort and led to several postponements, as patients with CKD often suffer from infections or require antibiotic prophylaxis due to their primary diseases, especially after transplantation or in advanced stages of CKD. In general, the efficacy of antibiotic prophylaxis prior to dental intervention is a controversial topic [31], but should be considered for the development of interventional trials in this field.

The identified organisms in this study are consistent with the work of former investigations [12, 13]. Comparing the prevalence, it has to be considered that most reports, among them the studies by Roberts et al., do not discuss the origin of the identified bacteria like Staphylococcus epidermidis or Staphylococcus hominis that are likely to be contaminants. This might result in an overestimation of the true prevalence of transient bacteremia in these studies [13]. Two of our isolates (20%, Streptococcus mitis and Streptococcus constellatus) belong to the large group of Viridans streptococci. Viridans streptococci represent the most frequently isolated group of bacteria with oral origin, resulting in a detection rate of up to 58% after dental procedures in children [12]. Propionibacterium acnes was detected in two blood cultures. Analogous to the interpretation by Hartzell et al. [24], contamination by skin bacteria seems highly probable despite the aseptic blood sampling technique that was performed according to the recommendations of the RKI [27‐29]. It can be speculated that the diversion of the first ml of each specimen would have further reduced the risk for contamination [32]. Even though a definite conclusion is not possible, most of the bacteria identified in the present study can also be found in the oral flora, supporting the assumption that those cases show indeed true transient bacteremia after toothbrushing [33, 34].

Concerning the detection method, sensitivity is the main issue. Blood culture bottles are the most commonly used medium [11, 12, 24‐26], remain the gold standard in clinical practice in German hospitals, and have been used in the preceding project to our study [35]. A more sensitive methods is lysis-filtration, being more time-consuming but allowing not only a more sensitive detection but also information about the bacterial load [36]. Furthermore, authors like Kinane et al. [37] found PCR-analysis to be more sensitive than culture methods (bacteria detection after toothbrushing in 13% vs. 3%) and Marín et al. [38] compared further molecular techniques like direct anaerobic culturing (DAC), lysis-centrifugation (LC), and quantitative PCR-analysis with detection via BACTEC^TM cultures. There is so far no generally accepted consensus on the optimal method to be used.

Another important potential influencing factor on the prevalence of bacteremia is the oral health status. To evaluate the oral situation of our participants, we used dental indices like dmft/DMFT, PBI, GI, and QHI (Table 1). Mean dmft/DMFT index was 0.6 ± 1 (median 0, range 0–3) similar to the very low prevalence of dental caries in German 12-year-olds (mean DMFT 0.7 [39]). In a preceding project [35], children with congenital heart diseases showed a higher prevalence of bacteremia after toothbrushing (21.4%) and also a greater caries prevalence with a mean DMFT of 4.6. Patients with CKD show lower caries prevalence, e.g., due to the alkaline oral pH that protects against acidogenic microflora associated with caries [40]. Whether or not caries influences the trespassing of bacteria into the blood stream is yet to be determined.

Mean PBI at baseline was 1.1 ± 0.7 (median 0.9, range 0.2–3.0) equivalent to a rather mild gingival inflammation [41]. Appropriately, GI according to Löe and Silness [42] was 1 ± 0.6 (median 1.0, range 0.2–3.0) at baseline. In addition, our participants demonstrated mild to moderate plaque accumulation as QHI [43] was 2.5 ± 1 (median 2.4, range 1.3–5.2) at baseline.

Gingivitis is one of the frequently reported soft tissue manifestation of CKD and is confirmed by the present baseline examinations of the dental indices. The increased plaque levels and gingivitis indices at baseline demonstrate inadequately performed oral hygiene measures prior to dental instruction and prophylaxis measures. Even in high-risk transplant patients, a demand-oriented preventive dental service can lead to an improvement in oral hygiene at home and thus improve the local inflammatory process in the oral cavity [40].

Concerning the relationship between oral health and risk for transient bacteremia, Silver et al. [44] described an increase from 16% prevalence in patients with a GI 0–0.75 to 68% in patients with moderately to severely inflamed gingiva (GI 2.26–3). Other investigators failed to prove a statistically significant impact of oral health on the trespassing of oral bacteria into the bloodstream [25, 45]. Nevertheless, Maharaj et al. [25] describe a difference in detection from 8% in GI 0.1–2.0 up to 16% in GI 2.1–3.0 and also an increase to 19% in patients with severe plaque accumulation which might be of clinical significance.

Regarding the mild to moderate gingival inflammation and plaque accumulation at baseline and even significantly lower indices after intensive treatment, the reported low prevalence of bacteremia is in line with former work on patients with similar oral conditions. Nevertheless, the given dental indices have to be interpreted with caution as they are calculated on the basis of averages and disguise the locally high plaque-induced gingivitis seen in our participants.

Another potential contributing factor to the prevalence of bacteremia is the grade of invasion caused by dental intervention [31]. In a study by Lockhart et al. [26], toothbrushing provided a 32% risk of developing bacteremia whereas tooth extraction without antibiotic prophylaxis provoked bacteremia in 80% of the cases. This finding is confirmed by other investigators such as Roberts et al. [12]. Even the difference between manual and electric toothbrushing is significant (48% vs. 78%), probably due to the more vigorous brushing with electric devices [11]. In the present study, we chose powered toothbrushing performed by three different investigators. Patients participating in our project were fully awake during the toothbrushing compared to several other studies where brushing was performed under general anesthesia prior to other dental interventions [11, 12]. Most studies on conscious participants have examined adults. This circumstance might have had an impact on the vigorousness and thoroughness during the toothbrushing and in consequence might be one of the reasons for the relatively low prevalence of bacteremia in our study. Furthermore, the conduct under general anesthesia facilitates the blood sampling itself as the vascular access is often easily available and the time interval of 30s between toothbrushing and blood sampling can be fulfilled more strictly.

Apart from transient bacteremia detected by blood culture, we also examined laboratory parameters, especially the CRP values of our participants, as markers for the chronic inflammation that constitutes one of the main factors for cardiovascular risk and subsequently for mortality. The fact that there is a significant association between gingivitis and chronic inflammation even in systemically healthy 15-year olds [46] emphasizes the need for interventional trials already in mild stages of oral inflammation. hsCRP values ranged from 0.2 to 8.4 mg/l with a mean value of 0.6 ± 0.9 mg/l [46] and were substantially lower than in our patient cohort (1.6 ± 1.7 mg/l, range 0.3–5.8 mg/l), which accentuates the special need for action in patients with CKD and gingivitis.

Referring to studies about the impact of non-surgical dental intervention in patients on systemic proinflammatory biomarkers such as CRP, data are scarce. Freitas et al. [19] showed a mean decrease of –0.2 mg/l in their meta-analysis. Fang et al. [47] examined 97 patients with stage 5 CKD and found a significant decrease in hsCRP 3 (–0.3 mg/l, p = 0.005) and 6 months (–0.4 mg/l, p < 0.001) after non-surgical periodontal therapy compared to the control group without treatment. Further studies on adults with CKD confirmed a possible link between periodontal therapy with improvements in oral health and decrease of hsCRP, increase in serum albumin and hemoglobin levels by improving erythropoietin responsiveness [48, 49]. In contrast, Wehmeyer et al. [18] failed to detect a significant improvement in serum albumin or hsIL-6 values, probably attributable to their study design and heterogeneous randomization, especially concerning participants affected by diabetes. Due to heterogeneity concerning the degree of CKD, severity of periodontal disease, periodontal treatment, and chosen outcome, comparison and final conclusion concerning the effect of dental intervention on inflammation and kidney function in CKD patients remain incomplete [19]. In our study, members of the IP group showed a statistically nonsignificant decrease in hsCRP by −0.2 ± 0.6 mg/l (available data n = 5) after receiving intensive prophylactic measures during the first intervention period. The decrease in the TAU group after receiving standard measures appears to be greater (−2.2 ± 2.5 mg/l, n = 6) due to the fact that the baseline value contains two outliers with singularly elevated CRP values (5.8 mg/l and 5.5 mg/l) without clinical signs of infection. Reasons for the elevation remain uncertain, but the singularity of the measurements is an argument for an acute event rather than mirroring chronic microinflammation. As recurrent infections are frequent in patients with CKD, interpreting changes in CRP is challenging.

There were also no significant changes in serum albumin, creatinine, and 25-OH-vitamin D associated with intensive dental care. The significant increase of 25-OH-vitamin D after TAU (mean change +9.8 ± 8.6 μg/l, p = 0.046) might also be attributable to changes in vitamin supplementation, kidney function with impact on the calciotropic hormone axis, or seasonal changes. All of our results have to be interpreted with caution given the limited sample size and data availability due to the exploratory character of the trial.

During 3 years of patient acquisition, several problems became obvious. We were not able to comply with the set interval of 10 ± 1 weeks as given by the study protocol in all of the cases. The mean interval between the visits was 122 ± 67 days (first intervention period) and 174 ± 101 days (second intervention period). The main reasons for this deviation were fluctuation in health as well as logistical difficulties on the part of the participants, especially given the additional effort required by an interdisciplinary clinical trial. Furthermore, not all laboratory parameters were available in all patients, resulting in a discontinuous monitoring throughout the study impeding conclusive evaluation.

Despite the fact that we were not able to show a statistically significant positive influence of intensive dental prophylaxis on transient bacteremia and chronic inflammation, there are implications that intensified dental treatment as applied by our study protocol does not only restore oral health but might also have a systemic positive impact. TAU as supported by the statutory health insurance is not sufficient to preserve or restore oral health in this cohort of chronically ill patients. It seems obvious to deduce that preventing periodontitis by early prophylaxis and improvement of dental care in this population is crucial, especially regarding the presence of chronic inflammation already in young age. Our trial can serve as a helpful pilot for further studies as it depicts valuable information on which potential pitfalls should be considered. Future analyses should treat a larger patient cohort with long-term follow-up and ensure continuous monitoring of laboratory parameters throughout the study period combined with bacteria detection by molecular methods. Implementing a needs-oriented intensive dental prophylaxis with regular appointments is a necessary approach in improving oral health and may also be a promising strategy concerning improvements in general health, systemic inflammation, and premature CVD.