Systemic Review on the association be-tween Periodontitis and Head and Neck cancer

Abstract

Background

The purpose of study is to estimate the association between periodontitis and risk of head and neck cancer by meta-Analysis.

Method

Literature was selected according to RPISMA guideline in PubMed, and Cochrane Library database. The association between periodontitis and the risk of having head and neck cancer was evaluated by meta-analysis using RevMan program. Random effect model was used and fixed effect model was analyzed as reference. Also, Subgroup analyses were done including covariates adjustment, study design, tumor site, ethnicity and type of assessment of periodontal disease. Publication bias were evaluated by Funnel plot using RevMan program. Eggers regression test was also performed by using comprehensive meta-analysis program to evaluate publication bias.

Result

Out of 13 studies selected, 14 results included in this meta-analysis. The asso-ciation between periodontitis and head and neck cancer was odds ratio (OR) was 2.35 with 95% confidence interval (95% CI) was between 1.69 and 3.26 for random effect model (p value < 0.00001) and OR was 1.27 with 95% CI was between 1.20 and 1.35 for fixed effect model (p value <0.00001). In Subgroup analysis, adjusted covariates was OR of 2.13 with 95% CI of 1.55 to 2.92. (p value < 0.0001) in random effect model. As for the result of study design, OR of case control studies was 2.51 with 95% CI of 1.77 to 3.57 in random effect model. Result of assessment method of periodontal disease, alveolar bone loss (ABL) was shown to be OR of 2.61 (95% CI [1.43, 4.79], random effect model) and clinical attachment loss (CAL) was OR of 3.66 (95% CI [0.67, 20.15], random effect model) and combining ABL and CAL shown to be OR of 2.68 (95% CI [1.62, 4.44], random effect model). In tumor site, the result of oral cavity was OR of 1.70 in random effect model (95% CI [1.24, 2.32]), head and neck was OR of 2.57 in random effect model (95% CI [1.60, 4.15]) and in ethnicity, the result of Europe was OR of 2.53 in fixed model (95% CI [1.82, 3.51]). As for the Eggers regression test, intercept val-ue was 3.44059 with p value of 0.00087 which shown that there is no publication bias.

Conclusion

The result of meta-analysis indicated that periodontitis could be associated with the risk of head and neck cancer. This kind of association could be different from covariates adjustment, study design, tumor site, ethnicity, and type of assessment of periodontal disease. To clarify this kind of link between periodontitis and risk of head and neck cancer, further studies is needed such as well-designed cross sectional studies or case control study with well-controlled sample size and type of assessment of periodontal disease, covariates adjustment, ethnicity, tumor site, etc