Enamel remineralisation protocol

For the in vitro enamel remineralisation model, 24 extracted human third molars were sectioned into enamel blocks and demineralised to produce artificial carious lesions, using a modified protocol to that described by White³⁴. Enamel blocks were each sectioned into an experimental half-block and control half-block. Each group of experimental half-blocks (n = 12) was suspended in the respective remineralisation solutions for 10 days at 37 °C with a change of solution every 48 hours. Subsequently, the half-blocks were paired with their control half-block and analysed for mineral content change using transverse microradiography (TMR) as described previously¹³.

The in situ clinical study had a randomised, controlled, double-blind cross-over study design. The University of Melbourne Human Research Ethics Committee approved the research conducted in this study (Approval Number 1441572) and the work was carried out in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki). The clinical trial registration number is ISRCTN 10532332 and date of registration is 23^rd November 2018. Artificially demineralised carious lesions were created on enamel blocks sectioned from human third molars and the blocks were each cut into experimental and control half-blocks as in the in vitro remineralisation model. Experimental half-blocks were inserted into custom-made palatal appliances to create a plaque retention site over the enamel lesions as described previously^10,23. Each appliance contained four enamel half-blocks with subsurface lesions, two on each side of the appliance^10,23. It was calculated that eight participants were required to provide the required statistical power (90%, p < 0.05) based on previous investigations using a similar cross-over in situ model^10,35,36. The required sample size was calculated using the G*Power Version 3.1 sample size package³⁷ and was based on a repeated measures analysis of variance with five levels, an effect size of 0.97, a correlation, ρ, between any pair of treatment means of 0.5 and a non-sphericity correction ε of 0.5. The effect size of 0.97 was based on detecting differences between ΔZd-ΔZr means of 70 (CPP-ACP + NaF and CPP-ACP + SnF₂) and a common standard deviation of 100 within groups. The non-sphericity correction adjusts for heterogeneity in the variances of the repeated measures. With a 5% significance level and a power of 90% at least six subjects were required. To allow for subject attrition eight subjects were recruited for the study.

Eight healthy adults (average age 43 ± 11 years; four males and four females) living in Melbourne, Australia with a fluoridated (0.9  ppm F), reticulated water supply participated in this double-blind, randomised, cross-over study. The participants were recruited from staff and students of the University of Melbourne and provided informed, written consent to participate in the study. Study inclusion criteria were: age 18–60 years; at least 22 natural teeth; unstimulated whole salivary flow rate of ≥0.2 ml/min and chewing gum-stimulated whole salivary flow rate ≥1.0 ml/min. Study exclusion criteria included: currently using antibiotics or medications that may affect salivary flow rates or a history of severe oral disease. The trial was conducted at the Royal Dental Hospital of Melbourne in 2014. Each participant wore their custom-made palatal appliance containing the four enamel half-blocks with subsurface lesions and rinsed with 5 mL of one of the five treatment solutions for one minute, four times each day for 14 consecutive days (treatment period) as described previously¹⁰. Participants were randomly assigned to each rinse. Each participant was assigned a number by the biostatistician (GDW) and randomisation was effected using a standard randomisation table for the five coded rinses. Participants were instructed to continue their normal dietary regime during the study and were given a toothbrush and sodium fluoride toothpaste (1450  ppm F) to brush their teeth twice a day. The intra-oral appliances were removed during oral hygiene procedures and were kept in a sealed humidified container. Participants were also instructed to clean their intra-oral appliances with a toothbrush and a fluoride-free denture paste supplied to them, taking care to avoid the attached enamel half-blocks. At the conclusion of the 14 days, participants rested from the study for one week (washout period) then began another treatment with another randomly assigned treatment solution. This was repeated until participants had rinsed with all five solutions. During the treatment periods subjects maintained a diary recording each rinse and duration of rinse. Following the treatment period, participants returned their appliance, empty rinse tubes and diary to the investigators and new enamel half-blocks were attached for the next treatment period. At the completion each experimental half-block was paired with its control half-block following the remineralisation period for analysis of mineral change using TMR and Sn and F levels using electron probe micro-analysis (EPMA). Researchers and participants were blind to the treatment code. A senior staff member held the treatment code securely which was only released after data collection and analysis.

Initial lesion depth (LDd), lesion depth change (LDd-LDr), initial mineral content (ΔZd), mineral content change (ΔZd-ΔZr), and percent remineralisation (%R) were measured from analysis of the demineralised and remineralised lesion mineral profiles from the TMR images as described previously by Cochrane et al.²³ For the in vitro data, a two-sample t-test was used to measure differences in lesion parameters (LDd, LDd-LDr, ΔZd, ΔZd-ΔZr and %R) between the two treatments. For the in situ data, the subject was the unit of analysis and the same lesion parameters were compared across the five treatments using analysis of covariance (ANCOVA). The primary outcome measure was integrated mineral gain/loss (ΔZd-ΔZr and %R) determined by TMR and the secondary outcome was Sn and F wt% enamel uptake measures determined using EPMA. Data were analysed for normality using Q-Q plots and the Shapiro-Wilk test and homogeneity of variance was tested using Levene’s test³⁸. Post hoc pairwise differences between treatments were performed on the estimated marginal means using the Sidak adjustment for multiple comparisons. The statistical significance was set at p < 0.05. SPSS software version 22 (IBM Corp. NY, USA) was used for all statistical tests.