This was an examiner-blind, randomized, stratified, two-treatment, parallel group, 24 week, clinical study in healthy adult volunteers with moderate gingivitis. The study was carried out at Salus Research Inc, Fort Wayne, Indiana, US, with the study protocol approved by an independent review board (U.S. Investigational Review Board, Miami, FL 33143) IRB number: U.S.IRB2013SRI/04. It was performed in accordance with the requirements specified in the Declaration of Helsinki and relevant local laws and regulations. All eligible subjects provided written informed consent before initiation of study procedures. There were no protocol amendments.
Subjects were ≥18 years old, in good general physical health with ≥20 natural teeth. At screening/baseline subjects were required to have moderate gingivitis (in the opinion of the clinical examiner and with a whole mouth MGI score between 1.75 and 2.30), a mean whole mouth PI score ≥1.5, and a minimum of 40 tooth surfaces where at least 50% was gradable for each clinical index (third molars, orthodontically banded/bonded, fully crowned, extensively restored, or grossly carious teeth were not included in the tooth count). Exclusion criteria included: pregnancy; breast feeding; allergy/intolerance to the study materials; current smokers or who had quit within the 6 months prior to the study; use of smokeless forms of tobacco; taking, or had taken, in the 14 days prior to the baseline visit, antibiotics, anti-inflammatory medication or a systemic medication that could affect gingival condition; participation in another clinical trial or use of an investigational oral care product within 30 days of baseline visit. Dentition exclusions included: current active caries or periodontitis that could compromise the study or oral health of the subject; restorations in a poor state of repair; partial dentures or orthodontic appliances; teeth bleaching within 12 weeks of screening; use of a chlorhexidine mouthwash within 14 days of baseline.
The primary objective of this study was to evaluate the gingivitis efficacy of a dentifrice containing 0.454% w/w SnF2, stabilized in a non-aqueous base, after a dental prophylaxis and 24 weeks twice daily brushing, as assessed by bleeding index (BI). Secondary objectives were to explore between treatment differences by BI at 12 weeks, and modified gingival (MGI), number of bleeding sites, and plaque index (PI) at 12 and 24 weeks. An exploratory objective was to explore the treatment response by evaluating the proportion of subjects moving between gingivitis severity category, as expressed by number of bleeding sites at the subject level ( <10%, >10%<30%, >30% bleeding sites).
Procedures and assessments
At the screening visit, subjects gave their written informed consent to participate in the study. Demographic, medical history and concomitant medications were recorded, followed by an oral examination and a gingival assessment that included a gross oral soft tissue (OST) examination, an oral hard tissue visual examination and assessment of dentition exclusions and gingival status. Within 28 days of the screening visit, eligible subjects returned to the site for the baseline visit with overnight plaque (subjects abstained from oral hygiene from 21:00 the night before the visit).
Subjects underwent a full OST examination and assessments of gingival inflammation (MGI), gingival bleeding (BI) and supra-gingival plaque (PI), carried out by the same examiner throughout the study to control for inter-examiner variability.
Eligible subjects (those with a mean MGI score between 1.75–2.30, a mean whole mouth PI score ³1.5 and who met all inclusion and exclusion criteria) were stratified based on gender and baseline mean whole mouth MGI scores (moderate gingivitis between 1.75 and 2.30, stratified byLow: ≤2.00/High >2.00 score) to ensure a balance in gingivitis across both treatment groups and randomized to one of two treatments according to a schedule provided by the Biostatistics Department of the study sponsor. All subjects received a dental prophylaxis using a conventional non-fluoride prophylaxis paste followed by flossing and removal of residual plaque by dental polishing to bring their teeth to zero plaque. This was checked by a second examiner to ensure complete removal of plaque.
Subjects were assigned one of two study dentifrices: a 0.454% SnF2 dentifrice containing 0.454% w/w SnF2 (Sensodyne Complete Protection, GSK Consumer Healthcare, Weybridge, UK; US marketed dentifrice) or a negative control dentifrice containing 1000 ppm fluoride as sodium monofluorophosphate (SMFP) (Colgate Cavity Protection, Colgate-Palmolive Co., New York, US; US marketed dentifrice). Colgate Cavity Protection was chosen as the negative control as it is considered representative of a regular toothpaste, and it is not indicated for antigingivitis. Subjects were instructed to apply a full ribbon of dentifrice to the head of a supplied manual toothbrush and brush their teeth in their usual manner for one timed minute twice daily (morning and evening). Study products were overwrapped to mask their identity as far as possible. The study examiner, study statistician, data management staff and other employees of GSK Consumer Healthcare or site staff who could have influenced study outcomes were blinded to product allocation.
Modified Gingival Index and Bleeding Index
Gingivitis was assessed using the MGI, BI and number of bleeding sites (as bleeding index score of 1 or 2). The MGI is a noninvasive evaluation of visual changes of severity and extent of gingivitis. MGI was assessed on the facial and lingual surfaces of two sites of each scorable tooth (papillae and margin/7-7 in each arch). Two scores were recorded bucally/labially (papilla and margin) and two scores lingually/palatally (papilla and margin). The MGI scoring system ranges from 0 (absence of inflammation) to 4 (severe inflammation), as described by Lobene et al. The BI assesses the number of bleeding points elicited on probing as a measure of gingival condition. Gingivae were air dried and then a ball-ended ‘community periodontal index of treatment needs probe’ was inserted into the gingival crevice to a depth of approximately 1 mm and ran around the tooth, gently stretching the epithelium. The BI was assessed on the facial and lingual gingival surfaces of each scorable tooth (7-7 in each arch). Three scores were recorded bucally/labially (distal, body, mesial sites) and three scores lingually/palatally. The BI scoring system was as follows: 0 = no bleeding after 30 seconds; 1 = bleeding upon probing after 30 seconds; 2 = immediate bleeding observed. The number of bleeding sites was calculated for each subject as the number of sites with a BI of 1 or 2 across all evaluable tooth sites .
The six-site modification of the Turesky Modification of the Quigley Hein Index (PI) was employed to assess plaque on all natural, gradable teeth . Plaque was first disclosed using a dye solution (Gum Red-Cote®, Sunstar Americas, Inc., Schaumburg, IL, US). For assessment, each tooth was divided into six areas including the mesiofacial, facial, distofacial, mesiolingual, lingual, and distolingual surfaces. Disclosed plaque was scored on a scale of 0 (no plaque) to 5 (plaque covering 2/3 or more of the crown of the tooth) .
All spontaneously-reported adverse events (AEs) or abnormalities in the OST examination were recorded from the screening visit until 5 days after the last study product administration. The relationship between the occurrence of each AE and the product was assessed by the investigator using clinical judgment and graded as mild, moderate or severe. Treatment emergent AEs were reported for the safety population (all randomized subjects who received the study treatment).
It was planned to screen enough subjects such that 100 would be randomized to treatment to ensure a total of 88 subjects (44 per treatment group) completed the week 24 assessment. With 44 subjects per treatment group the study was calculated to have 90% power to detect a difference between treatments of 0.07 units in BI after 24 weeks of treatment, assuming a standard deviation of 0.10, with a 0.05 two-sided significance level.
The primary population for the assessment of efficacy was the Intent-to-Treat (ITT) population, defined as those who received their study treatment and had at least one post-baseline efficacy measurement. The primary outcome variable was whole mouth mean BI at 24 weeks, calculated by taking the mean of the BI scores over all evaluable sites. Secondary efficacy variables were whole mouth mean BI at 12 weeks and, whole mouth mean MGI, bleeding sites, PI at all sites and at interproximal sites only, and BI and MGI by high/low MGI subgroup, at 12 and 24 weeks.
The BI, MGI, PI, and number of bleeding sites at 12 and 24 weeks were compared between treatments using analysis of covariance (ANCOVA). For all but PI, the model included factors for treatment group and gender, with baseline BI and MGI as covariates. The MGI stratification factor was not included in the model as the actual value was included as a covariate. In the analysis of number of bleeding sites, the baseline whole mouth mean BI score was included as a covariate. Model assumptions of normality and homogeneity of variance were tested and found not to be violated.
Subgroup analyses of the BI and MGI variables were performed based on the low and high MGI levels at stratification using ANCOVA with factors for treatment group, gender, MGI stratification and treatment by MGI stratification interaction, with baseline BI and MGI as covariates. For PI scores (overall and interproximal sites), the model included factors for treatment group, gender and MGI stratification, with the baseline PI score (overall or interproximal as appropriate) as a covariate.
For each subject, the percentage of bleeding sites at each time point was calculated as the number of sites with a bleeding score of 1 or 2, divided by the total number of bleeding sites assessed (multiplied by 100%). The percentage of bleeding sites was categorized into four groups: <10%, 10-20%, 20-30% and >30%.
All tests were two sided and performed at the 5% significance level under a null hypothesis of no difference between treatments. Model assumptions were checked and deemed to be acceptable.
Subjects who withdraw from the study early were included in the statistical analysis up to the point of when they withdraw. The drop-out rate over the 24-week study period was expected to be low (<10%) and therefore there was no provision for imputation for missing data.