Periodontal healing of mandibular second molar following the extraction of mandibular third molar. A Bayesian Network Meta-analysis for the identification of the best surgical technique.

Aims: To perform a systematic review and Bayesian network meta-analysis (NMA) of randomized controlled trials (RCTs) on effectiveness of surgical interventions for removal of mandibular third molar (M3M) on periodontal healing of adjacent mandibular second molar (M2M). Materials and methods: Medline, Cochrane and Embase databases were searched to identify RCTs up to 5 November 2014. Patients with M3Ms fully developed, unilaterally or bilaterally impacted, were considered. Outcomes were clinical attachment level gain (CALg) and probing pocket depth reduction (PPDr) with a follow-up longer or equal to 6 months. Results: 16 RCTs were included and categorized in 4 groups investigating: regenerative/grafting procedures (10 RCTs); flap design (3 RCTs); type of suturing (1 RCT); and periodontal care of M2M (2 RCTs). Eleven studies were at high risk of bias. Guided tissue regeneration (GTR) with resorbable (GTRr) and non-resorbable (GTRnr) membrane, and GTRr with anorganic xenograft (GTRr+AX) resulted showed the highest mean ranking for CALg (2.99; 2.80; and 2.29, respectively) and PPDr (2.83; 2.52; and 2.77, respectively). GTRr+AX showed the highest probability of being the best treatment both for CALg (Pr=45%) and PPDr (Pr=32%). Conclusions: GTR-based procedures with or without combined grafting therapies provide some adjunctive clinical benefit compared to standard non-regenerative/non-grafting procedures.