| Reference | Study type | Population | Exposure | Outcomes | Risk of bias «Additional comments for included studies...»2 |
|---|---|---|---|---|---|
| «Li R, Hou M, Yu L, ym. Association between periodo...»1 | SR/MA | Studies (13) conducted in 2005-2021 in USA, Europe, Asia and Brazil, oral squamous cell carcinoma patients and healthy controls | Periodontitis | OSCC prevalence (OR) | Low |
SR=systematic review; MA=meta-analysis; OSCC=oral squamous cell carcinoma; OR=odds ratio
| Reference | Comments |
|---|---|
| «Li R, Hou M, Yu L, ym. Association between periodo...»1 | Mainly case-control studies were included, also one cross sectional study and 2 cohort studies. Adjustment for tobacco and alcohol use was done in all case-controls studies that were included in meta-analysis. In large cohort studies adjustment for tobacco or alcohol use was not done. Heterogeneity was significant but explained by differences in severity of the periodontal disease. In analysis including only severe periodontitis patients, no heterogeneity was observed. Severity of periodontitis significantly affected results (a kind of "dose response") and association of mild periodontitis with oral cancer was not detected. Two large cohort studies detected association, but when combined with meta-analysis the association was found not significant. In cohort studies the majority of patients most probably have mild periodontitis, since the prevalence of severe periodontitis worldwide is approximately 20-30%. Included studies of Tezal et. al. had also patients with oropharyngeal cancers, but those studies were not analyzed in meta-analysis and are not therefore considered to cause indirectness. |
| Reference | Number of studies and number of patients | Follow-up time | Absolute number of events (%) I | Absolute number of events (%) C | Relative effect (95% CI) |
|---|---|---|---|---|---|
| Level of evidence: moderate. The quality of evidence is not downgraded due to heterogeneity because no heterogeneity is detected in the group of severe periodontitis patients. The quality of evidence is upgraded due to large effect and dose-response gradient. |
|||||
| «Li R, Hou M, Yu L, ym. Association between periodo...»1 | 12 studies, 245874 patients with periodontitis | OR 3.28 (1.87-5.74) | |||
| 5 studies, 2522 patients with mild periodontitis | OR 1.97 (0.73-5.37) | ||||
| 3 studies, 1258 patients with severe periodontitis | OR 4.23 (2.92-6.13) | ||||
| 2 cohort studies*, 228446 patients with periodontitis | cohort studies of patients RR 1.5 (0.93-2.42) | ||||
I=intervention; C=comparison; CI=confidence interval
The association of mild periodontitis (stages I-II) with OSCC was not statistically significant, but there was a strong association of severe periodontitis (stages III and IV) with OSCC.
*The two cohort studies «Chung SD, Tsai MC, Huang CC, ym. A population-base...»2, «Wen BW, Tsai CS, Lin CL, ym. Cancer risk among gin...»3 showed both statistically significant association with periodontitis and OSCC, but when they were combined with meta-analysis, the result was not statistically significant, most probably because of not overlapping confidence intervals. Studies had different comparison cohorts, and that may influence the results.
«Chung SD, Tsai MC, Huang CC, ym. A population-base...»2. The incidence rate of cancer during the 5-year follow-up period was 14.80 (95% CI 14.28–15.34) per 1,000 person-years in subjects with chronic periodontitis (CP). Cox proportional hazards regression revealed that the hazard ratio of cancer during the 5-year follow-up period for subjects with CP was 1.23 (95% CI 1.20–1.27) compared to that of the comparison cohort. Comparison cohort was defined periodontally healthy.
«Wen BW, Tsai CS, Lin CL, ym. Cancer risk among gin...»3. The incidence rate of cancer was 1.14 times higher in the study cohort than in the comparison cohort [confidence interval (CI) = 1.11-1.17]. The adjusted hazard ratio (HR) was 1.05 (95% CI = 1.00-1.11). A multivariable analysis showed that the periodontitis patients exhibited an elevated risk of developing oral cancer (adjusted HR = 1.79, 95% CI = 1.42-2.25). Study compared periodontitis cohort to gingivitis cohort.
This summary of evidence excluded reviews «Ma Y, Tuerxun N, Maimaitili G. Periodontitis and t...»4, «Gopinath D, Kunnath Menon R, K Veettil S, ym. Peri...»5 including studies with oropharyngeal cancer.