Candida species detection in patients with chronic periodontitis: A systematic review and meta‐analysis

Abstract Objectives To assess the Candida species occurrence rate and concentration in periodontal pockets in chronic periodontitis (CP) by meta‐analysis. Materials and Methods A search was performed of articles published between January 1, 2010, and October 1, 2020, in English and in Russian, in the electronic databases MEDLINE‐PubMed, Google Scholar, The Cochrane Library, ClinicalTrials.gov, Research Gate, eLIBRARY, and Cyberleninka (PROSPEROCRD42021234831). The odds ratio (OR), standardized mean difference (SMD), and 95% confidence interval (CI) were calculated using Review Manager 5.4.1 to compare the risk of CP when Candida spp. were detected in the gingival sulcus or periodontal pocket and to compare Candida spp. density counts in patients with CP and periodontally healthy patients. Results Twenty‐six studies were included in the systematic review and 11 were included in the meta‐analysis. The results showed that Candida spp. may increase the chance of CP development by 1.76 times (OR = 1.76; 95% CI = 1.04–2.99; Z = 2.10; p = .04; I 2 = 61%). More Candida spp. were found in patients with CP than in periodontally healthy patients (SMD = 1.58; 95% CI = 0.15−3.02; p = .03; I 2 = 98%). No data were found relating to the statistically significant influence of Candida glabrata, Candida krusei and Candida tropicalis on CP development. Conclusion We found that Candida albicans insignificantly increased the risk of CP development but, due to the heterogeneity of the included studies, further research is necessary to determine the exact role of Candida spp. in the development and course of the inflammatory periodontal diseases.

. The ability of the yeast to coexist with commensal and pathogenic bacteria is an important factor in their virulence which grants both microorganisms new characteristics and allows them to colonize new niches (Bamford et al., 2009;Bartnicka et al., 2019;Bernard et al., 2020;Sultan et al., 2018;Wu et al., 2015). It is worth highlighting that Candida spp. can adapt in different oral niches expressing different phenotypes and virulence factors according to the pH, oxygen, or polysaccharide availability. The periodontal pocket and gingival crevicular fluid are favorable media for C. albicans germination and hyphal tip growth. In comparison with yeast cells, Candida hyphae are more able to attach to host cells and penetrate the tissues (Bartnicka et al., 2019). C. аlbicans can interact with Streptococcus gordonii, Fusobacterium nucleatum, P. gingivalis, and A. actinomycetemcomitans, to form mixed biofilms, which makes C. аlbicans an active participant in the inflammatory-destructive process in periodontal diseases (Bartnicka et al., 2019;Kornman, 2005).
The interaction between C. albicans and microorganisms seems to be complex and has been incompletely studied while the role of C.
albicans in the CP pathogenesis is also complex. They secrete proteinases that release toxic or antigenic agents which in turn can increase tissue inflammation and activate an immune response similar to that of P. gingivalis proteases (Sardi et al., 2010). Candida spp. are also able to secrete phospholipases which facilitate its adhesion to the tissues and destroy cellular membranes thereby promoting cytolysis (De-La-Torre et al., 2018).
Scaling and root planning is a gold standard in CP treatment.
However, in some patients, periodontitis is resistant to conventional treatment and requires systemic antibiotic therapy. This may promote active yeast growth which adversely affects the course of the CP (De-La-Torre et al., 2018;Sardi, Duque et al., 2011). Thus, refractory periodontitis may be linked to uncontrolled growth and replacement by the yeast-like fungi in a periodontal pocket.
A systematic analysis of the literature and meta-analysis of data to compare the detection rate and density counts of Candida spp in patients with CP and with periodontally healthy patients can explain their role in the origin and development of CP.
In this review, we wanted to look in detail at the occurrence rate of yeast-like fungi in CP and assess their density counts in periodontal pockets by analyzing the clinical studies published over the last 10 years.

| Review questions
A systematic review and meta-analysis was performed to compare the occurrence rate of Candida spp. in a periodontal pocket or a gingival sulcus in patients with CP and periodontally healthy patients, to determine the potential role of the yeast-like fungi in the development and course of CP.
2.2 | Search strategy, study selection, and data extraction (periodontal disease), (Candida) AND (periodontal pocket), (oral fungal-bacterial biofilm), (Candida-associated periodontitis). We activated the filter "human" in the PubMed database. Reference bibliographic lists of the retrieved found publications for full-text screening were reviewed and potentially significant studies were manually selected. No language, country, observation duration, and subject characteristics (race, age, and sex) were introduced.

| Study selection
The search results were downloaded to a bibliographic database (Mendeley Reference Manager) to facilitate duplicate removal and cross-reference checks.
Two authors (E. S., E. T.) independently screened the titles and abstracts of the entries identified in the search of all articles against the below pre-defined inclusion and exclusion criteria.
Inclusion criteria were as follows: • randomized and nonrandomized clinical studies, observational studies (cross-sectional, case−control); • patients over 18 years old, independent of race, nationality, sex; • patients with CP without periodontal treatment, antibiotic, and antifungal therapy in the previous 6 months; • Candida spp. isolation from the content of a periodontal pocket, subgingival biofilm, subgingival dental plaque, gingival sulcus; • determination of Candida spp. detection rate; • Candida spp. density counts in periodontal pockets; • full-text article availability; • English and Russian-language publications.
Exclusion criteria were as follows: • reviews, meta-analyses, clinical case studies, and reports; • in vitro studies; • studies with incomplete initial data; • studies where only the abstract was available; • studies where Candida spp. were detected in the saliva and on the oral mucosa; • studies on oral candidiasis; • the articles reviewed another pathology but not chronic generalized periodontitis; • patient was a child (under 18 years old); • the study included only subjects with comorbidity and immunodeficiency conditions. Next, the full-text version of all studies that potentially met the eligibility criteria or for which there was insufficient information in the title and abstract to make a decision were obtained.
Any disagreements between reviewers were solved by a joint discussion.

| Data extraction
Every author independently analyzed the publications and extracted data which was entered into the predetermined information table. Any disagreements were solved by a joint discussion. The following data were recorded from all the selected studies: name of the first author, publication year, country of the study, study design, studied population characteristics, sample size, number of subjects in the studied and control groups, the technique for sampling collection from the periodontal pocket or gingival sulcus, Candida spp. identification technique, study result expressed in Candida spp. detection rate or density count.
Control and experimental groups were respectively defined as a group of subjects without signs of periodontal inflammation and a group of patients with CP independent of the disease severity. As per the studies where groups were formed according to the severity of CP, the study subjects were combined in one group. In the assessment of the Candida spp. density count, the mean was calculated for groups with CP of different severity, when necessary, and a standard deviation parameter was calculated according to the average standard error value.

| Evaluation of the methodological quality of the studies
Two authors independently evaluated the quality and evidence level of every study using the Russian-language version of the Newcastle −Ottawa scale questionnaire to assess the risk of systemic bias in nonrandomized comparative studies (Rebrova & Fediaeva, 2016).

| Statistical analysis
Statistical analysis of the results was performed using Review Manager (RevMan) 5.4.1 software. An odds ratio (OR) and 95% confidence interval (CI) were calculated for dichotomic (binary) data while the standardized mean difference (SMD) and 95% CI were calculated for continuous data. Studies, where the above-mentioned data could not be calculated, were not included in the statistical analysis. Results of single studies and summarized data were presented in a forest plot. The heterogeneity of the studies was assessed by χ 2 and I 2 . The random effects model was used when p < .10 or I 2 > 50% determined significant heterogeneity across all articles. Otherwise, the fixed effects model was used. The analysis of sensitivity to the combined result stability was performed to examine the source of heterogeneity. The Candida spp. detection rate and individual Candida spp. detection rate were analyzed in groups of patients with clinically healthy periodontium and CP.

| Assessment of publication bias
The publication bias was assessed by funnel plots.

| Study selection
The article selection process is demonstrated in Figure 1. A total of 1709 publications were initially selected by name, abstract, and publication date. One hundred and twenty-three studies were excluded as duplicates. Six hundred and fifty-eight more articles were excluded for nonconformance of the title and abstract to the inclusion criteria, and 72 were excluded for the absence of the full text. Sixty-eight out of the remaining 126 full-text articles were excluded for nonconformance to the article selection criteria. Disagreements regarding inclusion or exclusion were settled by discussion. Finally, 26 publications were selected for the systematic review by the selection criteria. Detailed analysis of the presented data revealed that 14 articles were without a control group of patients with a clinically healthy periodontium which did not allow a calculation of the OR for the meta-analysis. Thus, the metaanalysis included 11 articles.
F I G U R E 1 Flow diagram of study selection SLAZHNEVA ET AL. | 1357 3.2 | Assessment of the methodological quality of the studies included in the meta-analysis An assessment of the methodological quality of the studies by the Russian-language version of the questionnaire to assess the risk of systematic bias in nonrandomized comparative studies (Newcastle −Ottawa scale) showed that two studies (Brusca et al., 2014; Risovannaya & Lalieva, 2019) had a high risk of systematic bias-5 points on the scale, 9 (Canabarro et al., 2013;De-La-Torre et al., 2018;Komleva et al., 2010;Krishnan et al., 2020;Matić Petrović et al., 2015Melekhov et al., 2013;Colombo et al., 2016;Volchenkova et al., 2017) studies had an average risk of systematic bias-6−7 points on the scale. Table 1 shows the data summary of the studies included in the systematic review and meta-analysis.

| Included study characteristics
The age of the subjects was from 18 to 80 years old. Three studies (Brusca et al., 2014;Krishnan et al., 2020;Melekhov et al., 2013) included only male subjects while subjects from both sexes participated in the other studies. A total of 2717 subjects were examined. Four hundred and sixty-three of them had a clinically healthy periodontium while 2254 had CP of differing severity.
C. krusei was detected in: • 5 (7.2%) patients with CP The study is included in the meta-analysis.
The result differed from the previous conclusion (Figure 3a) after the study by Krishnan et al. (2020) was excluded. There was, however, no grounds to exclude the study from the meta-analysis.
3.4.2.2 | C. glabrata No data was received about a statistically significant influence of C.

| C. krusei
No data were received about the statistically significant impact of C.

| DISCUSSION
Bacteria P. gingivalis, T. forsythia, T. denticola, and A. actinomycetemcomitans are pathogens which are most frequently associated with inflammatory periodontal diseases (Socransky et al., 1998 are of increasing interest, for example, C. glabrata, C. tropicalis, C. krusei, and C. dubliniensis, which have recently been isolated from the oral cavity not only in immunocompromised, but also in healthy subjects (Matic Petrovic et al., 2019;Quindós et al., 2018).
The present work represents a quantitative analysis, which combines the results of independent studies with different designs. The metaanalysis of 11 selected studies demonstrated that in CP patient group, the detection rate of Candida spp. was statistically significantly higher and increased the risk of CP development by 1.76 times compared to the group of subjects with clinically healthy periodontium. Statistically significant data also confirmed that a higher density of Candida spp.
was detected in periodontal pockets of CP patients than in subjects with clinically healthy periodontium. However, these data should be interpreted with care as a high heterogeneity across the studies was observed; publication bias was present across the studies where Candida spp. were more frequently detected and their (Candida spp.) density was higher in advanced periodontitis and the studies had a moderate or high risk of systemic error. There are few studies of species such as C. glabrata, C. krusei, C. tropicalis, C. parapsilosis. For example, our meta-analysis detected only two studies that corresponded to the inclusion criteria and that compared the detection rate of C. glabrata, C. krusei, C. tropicalis in periodontal pocket samples and crevicular samples in CP patients and subjects with clinically healthy periodontium respectively (Krishnan et al., 2020;Volchenkova et al., 2017). The detection rate of these species was similar in both groups. Colonization of periodontal pockets by these microorganisms did not necessarily certify their activity in the pathogenesis of periodontitis. They can be transient members

| CONCLUSIONS
In conclusion, we would like to emphasize that our meta-analysis is one of the first that critically assesses the detection rate and density

ACKNOWLEDGMENTS
This study was supported by the Russian Association of Periodontology.

CONFLICT OF INTEREST
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
Data are available on request from the authors. The data that support the findings of this study are available from the corresponding author upon reasonable request.