Diederik Hentenaar

114 Chapter 5 were cleaned with either local application of 35% phosphoric acid gel (pH 1) for 1 minute (Temrex gel, Temrex, Freeport, NY, USA) (test group) or by rinsing with an abundant amount of sterile saline for 1 minute (control group). Care was taken to apply the phosphoric etching gel precisely on the implant surface using a syringe with a small tip. During one minute the etching gel was continuously rubbed on to the implant surface with a small brush. In both groups, the intervention continued with rinsing of the implant surface with an abundant amount of sterile saline for 1 minute. Angular bony defects were eliminated and bone was recontoured using a rotating round bur under saline irrigation. Mucosal flaps were apically positioned and firmly sutured (Vicryl Plus 3-0; Ethicon Inc., Somerville, NJ, USA) and suprastructures were re-positioned. For both control and test group, surgery was followed by 2 weeks of mouth rinsing with 0.12% CHX + 0.05% CPC without alcohol two times daily for 30 s. Sutures were removed after 2 weeks. Follow-up visits were scheduled after 3 (T3) months. Patients were all surgically treated by one experienced oral- and maxillofacial surgeon (GR). Outcomes Primary outcome variable The primary outcome variable was the difference in anaerobic bacterial load of the implant surface before and after mechanical and chemical debridement and decontamination. After flap deflection and granulation tissue removal a sample was obtained from the implant surface by rubbing a sterilized brush (Microbrush® International, Grafton, WI, USA) across the implant surface (Tpre). A second sample was obtained after mechanical debridement, decontamination of the implant surface with the test or control substance and subsequent rinsing with sterile saline (Tpost). After sampling the top part of the brush was cut off and collected in a vial containing reduced transport fluid (Syed & Loesche, 1972). From every implant presenting peri- implantitis separate samples were obtained. All microbiological samples were processed within 24 h (Van Winkelhoff et al. 1985). The total anaerobic bacterial load and the presence and numbers of the periodontal pathogens (Zambon 1996) Aggregatibacter actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, Tannerella forsythia, Fusobacterium nucleatum, Parvimonas micra and Campylobacter rectus were determined by laboratory technicians who were blind to treatment allocation. Secondary outcome variables Secondary outcome variables were percentage of sites with bleeding on probing (% sites BoP), percentage of sites with suppuration on probing (% sites SoP), mean probing pocket depth (mean PPD) and microbial composition of the peri-implant sulcus. Measurements were performed before (pre) treatment (baseline, T 0 ) and at 3 months (T 3 ) after surgery by one and the same examiner (DH) who was blind to