Clinical evidence in the treatment of white spot lesions following fixed orthodontic therapy: a meta-analysis

45 © Australian Society of Orthodontists Inc. 2018 Objective: This systematic review aims to determine the most effective method of treatment to remineralise post-orthodontic white spot lesions (WSLs). Method: Six databases were accessed and searched for articles. Screening and selection were conducted according to the PRISMA guidelines using predetermined inclusion and exclusion criteria. Two reviewers independently assessed and extracted identified studies and relevance disagreement was resolved through consensus. Experimental studies were included that involved (i) patients of any age who had WSLs after the removal of fixed appliances, (ii) any treatment to remineralise the WSLs compared with no treatment or a placebo, and (iii) measurement of the changes in enamel mineralisation status after treatment. Eligible articles were assessed for internal bias and underwent narrative synthesis. A meta-analysis using random-effects modelling was performed to calculate a pooled estimate and assess between-study variability using Cochran’s Q and I2. Results: The nine articles included in this review were found to have a medium or high risk of bias. The qualitative assessment provided contrasting results between studies. The meta-analysis showed both CPP-ACP – pooled d of -0.28 (N = 5 studies; 95% CI = -0.48-0.07) – and fluoride – pooled d of -0.25 (N = 4 studies; 95% CI = -0.48-0.02) – to generate improvement in the enamel mineralisation status, with CPP-ACP producing more consistent results compared with fluoride. Conclusions: The meta-analysis found that CPP-ACP and fluoride were effective in reducing post-orthodontic WSLs. Due to the heterogeneity of the included studies with regard to the fluoride concentrations and mode of delivery, the current meta-analysis could not accurately establish which remineralising agent, CPP-APP or fluoride, is more effective. Further high quality studies of long-term duration are required to determine best clinical practice. (Aust Orthod J 2018; 34: 45-60)


Introduction
White spot lesions (WSLs) are localised porosities of subsurface enamel caused by demineralisation. 1 Light striking a hypocalcified enamel surface scatters, resulting in an opaque white appearance 2 that visually diagnoses WSLs in clinical practice. For research purposes, various indices have been used to quantify the severity of WSLs. 3 Quantitative lightinduced fluorescence (QLF) is a commonly used and highly sensitive diagnostic tool that utilises the relationship between the mineral content of a tooth and the amount of fluorescence emitted from enamel after being exposed to high-intensity blue light. To illustrate, a low QLF reading would be produced from a demineralised tooth structure since it emits less fluorescence and a greater lesion depth is reflected by a higher loss of fluorescence (∆F). QLF also allows researchers to precisely measure the area of the WSLs (A). Similarly, laser fluorescence (Diagnodent) can be used to determine the depth of a lesion by measuring the loss of dental fluorescence. 4 Eng Seng Yeoh: eng.yeoh@uq.net.au; Tam  WSLs are considered the most common iatrogenic complication of orthodontic treatment as the introduction of fixed appliances can unfavourably alter the oral environment. 5 Prevalence studies carried out using QLF reported that WSLs were found in 97% of post-orthodontic cases. 6 Orthodontic brackets, ligatures and bands physically obstruct effective cleaning in the area around these appliances and reduce the natural self-cleansing effect, giving rise to prolonged plaque accumulation and retention. 7,8 Consequently, the imbalance between the processes of demineralisation and remineralisation of enamel results in the creation of opaque hypocalcified lesions. 2 While demineralisation usually decreases after the removal of fixed orthodontic appliances, some lesions persist and result in compromised aesthetics. 1 In severe cases, WSLs may progress to cavitated carious lesions and require invasive restorative intervention. 9 Therefore, effective treatment following fixed orthodontic therapy is necessary to reverse and remineralise these potentially disastrous and unsightly lesions.
Many options have been attempted and described to treat WSLs. For example, fluoride supplements in the form of varnishes, dentifrices and mouthwashes have been routinely used. 9 While the strengthening capacity of fluoride on enamel and its effectiveness in preventing WSLs is widely accepted, the current literature has limited evidence to support the efficacy of fluoride in remineralising post-orthodontic WSLs. 10 In recent years, attention has been directed to calcium phosphate-based remineralisation agents. The casein phosphopeptide-amorphous calcium phosphate (CPP-ACP) containing agent functions by concentrating calcium and phosphate substrates in saliva, which enhances its buffering capacity. 11 As a result, the remineralisation process is enhanced whereas the demineralisation process is suppressed. 12 Previous in vitro studies and clinical trials that investigated the effectiveness of CPP-ACP in reducing demineralisation around orthodontic brackets have shown favourable results. 13,14 However, there is a lack of clinical consensus regarding the effectiveness of remineralising agents in treating post-orthodontic WSLs. A systematic review published in 2012 comparing the effectiveness of CPP-ACP against fluoride in reducing post-orthodontic WSLs produced inconclusive findings and a lack of robust quantitative results. The authors also reported that most of the available studies had methodological problems. 15 Hence, the present review aims to (1) examine the effectiveness of remineralising agents in reducing post-orthodontic WSLs and (2) produce a meta-analysis of the remineralising agents.

Search strategy
Six databases (PubMed, EMBASE CINAHL, Cochrane, Scopus and Web of Science) were searched using a combination of key search terms relating to 'orthodontic', 'white spot lesions', 'CPP-ACP', and 'fluoride' (see Appendix I for the detailed searched strategy). Articles were limited to those published between 2005 to 2016 and written in English.
A systematic screening of the searched articles was conducted using the PRISMA guidelines. 16 Following the initial search, duplicate articles were removed and the title and abstract of the resulting articles were assessed for relevance by both reviewers (TL and EY). The full texts of the remaining publications were analysed and articles that did not meet the inclusion criteria were excluded. The reference lists of all the full text papers were scanned to identify related studies.
The criteria for article inclusion in the present review were patients of any age who had WSLs after the removal of fixed appliance, those receiving treatment to remineralise the WSLs, and studies that compared no treatment with a placebo or alternative treatment. Moreover, the outcome of interest was the change in the status of enamel remineralisation, assessed through visual examination, photographs, quantitative light induced fluorescence or Diagnodent. Studies were also required to be an in vivo randomised controlled trial investigating treatment of white spot lesions following completed fixed orthodontics. Studies with artificial white spot lesions or those not comparing an intervention were excluded.

Quality assessment
To assess potential bias in the methodology of the selected studies, the Cochrane Collaboration's tool for bias appraisal was applied. 17 The two reviewers independently assessed each study for low, unclear or high risk of bias in terms of the randomisation method and blinding of randomisation, blinding of participants and examiners, blinding of outcome, attrition of data, bias in the reporting of results and other possible sources of bias. The studies were classified as low risk of bias (low risk of bias for all key domains), medium risk of bias (low or unclear risk of bias for all key domains) or high risk of bias (high risk of bias for one or more key domains) to summarise the assessment results. In cases of differing opinions between the reviewers, the concerns were discussed with an additional reviewer (HH) to reach a consensus.

Data extraction
Data from the eligible studies were extracted independently by the reviewers (TL and EY) and disagreement was resolved through consensus and after discussion with a third reviewer (HH). Information obtained included the study characteristics (year of publication, type of study) and sample characteristics (sample size, inclusion criteria, follow-up duration), as well as intervention and study results (including the method of assessment). Effect estimates such as regression coefficients were obtained from each study. In the absence of effect estimates, standardised mean difference was derived using the sample size, mean and standard deviation through Practical Meta-analysis Effect Size Calculator. 18

Meta-analysis
Only studies that compared the remineralising treatment to a placebo, normal oral hygiene regimen or no treatment were considered in the meta-analysis. The meta-analysis was compiled using a qualityeffects model instead of random-effects model as the former considers the quality score as it pooled the effect estimates of each study. 19 Between-study heterogeneity was assessed using Cochran's Q and I 2 statistical method. Separate meta-analyses were completed according to the type of intervention (CPP-ACP and fluoridation). Subgroup analysis per type of outcome (fluorescence score, lesion area and clinical index score) and sensitivity analyses were conducted to explore the changes in pooled estimates and the level of heterogeneity.

Study characteristics
A total of 708 articles were retrieved from the six databases, as shown in Figure 1. After the removal of duplicates, 435 titles and abstracts were examined for relevance and 24 articles were selected. After evaluating the full texts, nine studies were finally considered eligible based on the selection criteria and subsequently reviewed. The description of each study is provided in Appendix II. In total, 15 articles [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34] were excluded, six 20-25 because the experimental design used artificial lesions in vitro, two 26,27 because there was no comparison of a remineralising intervention against a control. The participants of one study 28 had not undergone orthodontic treatment, one study 29 did not include statistical data in the results section and was therefore excluded since a meta-analysis or comparison of results could not be performed. One article was written in Chinese with an English version unable to be located 30 and four articles 31-34 were reviews.
After the study period, the average lesion area was decreased by 58% in the CPP-ACP group and by 26% in the control group, which was significantly different to baseline in the intervention group (p < 0.05).
The difference between two groups were not statistically significant (p = 0.06).
Regarding the clinical score, there were no statistical significant difference between the two groups. There were statistically significant differences between the mean DD readings of the two groups at the threemonth (p < 0.05) and at the six-month follow-up visits (p < 0.01).
There was a significant decrease in the mean DD reading in group 1 during the first three months (p < 0.001) and a further decrease after six months. The decrease in mean DD reading between the third and the six months were not statistically significant (p > 0.05 The sample size of the included studies ranged from 24 to 240. All of the participants of the randomised control trials were recruited from the orthodontic department of the respective universities and underwent fixed orthodontic treatment immediately before the trials.

Fluoride versus inactive control
In a study that compared the efficacy of 5% fluoride varnish against 5% acidulated fluoride film and a control paste, the authors reported that the improvement in ∆Q (∆F × A) was highest in the fluoride varnish group, followed by the fluoride film group and the control group, respectively. The difference between the three groups was statistically significant (p < 0.05). 42 Bailey   (p < 0.05) improvement in fluorescence loss in both the intervention and control groups was reported, but no difference between the two groups. There was also no clinical or statistical significance in lesion area (A) over time. 39

CPP-ACP versus fluoride
Andersson et al. compared the effect of CPP-ACP against a combination of fluoridated toothpaste and 0.05% fluoridated mouthrinse for 12 months. Diagnodent and visual scoring were used to evaluate the outcomes. At the end of the treatment, the Diagnodent readings showed statistically significant improvements in the test and control groups, although the difference between the two groups was not statistically significant. However, visual scoring showed a higher percentage of lesion regression in the test group compared with the control group (64% vs 23%), the difference noted as statistically significant (p < 0.01). 36 Bröchner et al. compared the effect of CPP-ACP against normal fluoridated toothpaste over four weeks, no clinical or statistical differences were found using QLF and clinical scoring. While there was a 58% decrease in the average lesion area (A) in the CPP-ACP group and 26% in the control group, the difference was not statistically significant (p = 0.06). An evaluation using clinical scoring also reported no statistically significant differences between the reduction of WSLs in the two groups. 40

CPP-ACP versus fluoride versus non-active control
Huang et al. conducted a study involving two intervention groups, CPP-ACP and 5% fluoride varnish, which were compared with a control group in which the respondents were instructed to brush with 1100 ppm fluoridated toothpaste. A linear regression model showed no statistically significant difference between the improvement in the three groups. 43 In contrast, Akin et al. reported that the CPP-ACP group showed a higher success rate in the treatment of WSLs compared with the 0.025% fluoride rinse group and the inactive control group (p < 0.01). 35

Quality assessment
The outcome of quality assessment of the studies is shown in Table 1. Overall, five studies 35,[37][38][39]41 were evaluated to have a high risk of bias and four 36,40,42,43 assessed to be of medium risk. Baeshen et al. 37 and Du et al. 41 were graded as a high risk of bias in allocation concealment as a randomisation list/table was used, while Akin et al. 35 grouped participants according to the date of debonding, which may have led to selection bias. Beerens et al. 39 was graded as high risk and Bröchner et al. 40 was graded unclear in attrition bias due to the high and moderately high loss of follow-up data, respectively. Bailey et al. 38 did not report the pre-specified QLF result and also had a high risk of bias due to funding of the study provided by the manufacturer of the product.
Surprisingly, CPP-ACP estimates improved in statistical significance across all types of WSL

Selection bias
Random sequence generation

Performance bias
Blinding of participants and personnel. Assessments should be made for each main outcome (or class of outcomes).

Detection bias
Blinding of outcome assessment. Assessments should be made for each main outcome (or class of outcomes).

Attrition bias
Incomplete outcome data. Assessments should be made for each main outcome (or class of outcomes).

Reporting bias
Selective reporting

Other bias
Other sources of bias   37 which also measured the clinical index score, as well as the study by Akin et al. 35 which had poor randomisation, improved the estimates for fluoride treatment and diminished the score for CPP-ACP (see Table II).

Discussion
The present systematic review incorporating a meta-analysis of the pooled data focused on determining the most effective method of remineralising post-orthodontic WSLs. Alternative treatment methods such as micro-abrasion and resin infiltration were not considered. Although there were in vitro studies 13,14,44 that reported positive and significant results, these were excluded as they did not adequately replicate the oral environment. Additionally, many confounding factors such as patient compliance, diet and salivary buffering capacity can also affect the clinical result of WSLs. 41,45,46 Hence, only randomised controlled trials were considered because the randomisation aimed to distribute any confounders equally between the intervention and control groups.
It is widely accepted that the introduction of fluoride as an anti-cariogenic agent was one of the most important events in the history of dentistry. 47 In addition to the enhancement of remineralisation and inhibition of demineralisation, fluoride has also been proven to have limited anti-microbial activity against bacteria in the oral biofilm. 48 While most of the included studies 35,37,41,42 in the review reported that additional fluoride has significant clinical benefit in reducing post-orthodontic WSLs, Huang et al. 43 reported that fluoride varnish was not more effective compared with normal oral hygiene instruction and fluoridated toothpaste. Huang et al. also argued that the use of high-concentration fluoride varnish may prevent the infiltration of calcium and phosphate into the deeper layer of the enamel, therefore inhibiting deeper remineralisation. 43 Akin et al. suggested that low-concentration fluoride rinse was more effective in small compared with large lesions due to better uptake of fluoride in shallow lesions. 35 Currently, in vitro 49,50 and in vivo studies 51,52 have demonstrated and validated the anti-cariogenic effects of CPP-ACP. Shen et al. 53 reported that CPP-ACP had superior efficacy to 5000 ppm fluoride in reducing the depth of WSLs. However, there is insufficient evidence to establish a clinical benefit of CPP-ACP in managing post-orthodontic WSLs. In the present review, two RCTs compared the efficacy of CPP-ACP against an inactive control paste and reported a benefit of CPP-ACP that was clinically but not statistically significant. 38,39 A variation in oral hygiene habits between the respondents may be considered as a confounding factor. Two of the included studies compared CPP-ACP against fluoridated toothpaste (± 0.025% fluoridated mouthrinse). 36,40 One 36 reported significant clinical improvement in the CPP-ACP group whereas the other 40 reported no significant result. Interestingly, two of the included studies compared CPP-ACP and fluoride against a normal oral hygiene routine but reported completely different results. Akin et al. 35 demonstrated that CPP-ACP was superior compared with fluoride and fluoride produced superior results to the control group in reducing post-orthodontic WSLs. In contrast, Huang et al. 43 reported no difference in effectiveness between the three groups.
Importantly, it was observed that 'no treatment' in most of the included studies seemed to result in lesion regression. 35,[37][38][39][41][42][43] As the 'no treatment' groups were not restrained from tooth brushing with fluoridated toothpaste, it can be postulated that fluoridated toothpaste may be responsible for the improvement. This is supported by the study carried out by Argawal et al., which concluded that fluoridated toothpaste can cause a significant reduction of post-orthodontic WSLs. 29 This also implies that the therapeutic benefits of the remineralising agents studied may be masked by the therapeutic effects of the daily oral hygiene regimen.
In general, the studies analysed in the present review yielded different results regarding the effectiveness of the remineralising agents on WSL outcomes. However, the meta-analysis indicated that CPP-ACP and fluoride are both effective in reducing postorthodontic WSLs. Studies concerning CPP-ACP demonstrated more consistent results (lower I 2 and Q). This finding could be related to the consistent dosage in CPP-ACP cream compared with the various concentrations of fluoride products available. The present analysis also inferred a possible clinical significance despite the statistical disadvantage of the results (e.g., low statistical power). 54 This was clearly observed in the improvement of the effect size of CPP-ACP after pooling individual study estimates. It was not possible to accurately establish which remineralising agent, CPP-ACP or fluoride, is more effective due to the heterogeneity in the included studies with regards to the concentration of fluoride applied and the mode of fluoride delivery.
Significantly, the inability to include the study by Huang et al. 43 may undermine the validity of the meta-analysis. However, it was methodologically impossible to include this study due to the lack of information regarding the mean difference. All of the other included studies [35][36][37][38][39][40][41][42] reported the pre-and posttreatment value of the enamel mineralisation status while Huang et al. 43 only reported the improvement score after treatment.
Significant heterogeneity was found in the instruments used to measure the improvement of WSLs, as each of the instruments had different sensitivity and specificity, which may have led to error when comparing the effectiveness of the treatment methods. Many studies have advocated that QLF is precise and consistent in quantifying and monitoring the mineral content and size of WSLs. [55][56][57] While the Diagnodent device can be useful in quantifying WSLs, it was suggested that its statistically significant difference may not coincide with a clinically significant difference. 58 Furthermore, stain and calculus can also affect the Diagnodent readings. 59 Visual assessment is also relevant when measuring the WSLs as this is the most common approach in clinical practice. Clinical indices provide a means to quantify WSLs but lack sensitivity in detecting small change. Photographic techniques, as described by Huang et al. 43 and Akin et al., 35 allow quantification of the area of WSLs. However, bias may be introduced with inconsistent angles and lighting. According to Bröchner et al., a combination of QLF and visual assessment could be beneficial in future studies. 40 The disparity related to the instruments used was found in the subgroup analysis of fluoride treatment but not the CPP-ACP analysis. Studies with fluorescence as the outcome, measured using QLF and Diagnodent, showed a higher pooled estimate compared with the lesion area score measured by imaging software.

Strength and limitations
Having a robust statistical analysis that clarifies the findings of the systematic review is the most significant strength of the present review. However, there are several limitations as the included studies show clinical heterogeneity in regards to the type and dosage of treatment and outcome measurement tools. This prevents direct comparison of the effectiveness of different treatment methods. The exclusion of non-English papers also introduced a level of bias by preventing an assessment of results from other countries. 60 The small number of eligible studies limited the ability to examine the relationship between the effect of treatment and the severity of WSLs at this baseline. Limited data also prevented the investigation of the aggregate dosage-response relationship of each treatment method and the effect modification across different follow-up periods. The inability to include the study by Huang et al. 43 may have also undermined the validity of the meta-analysis.

Recommendation
From the present review, it is apparent that practising good oral hygiene plays an important role in reducing post-orthodontic WSLs. Therefore, it is the clinician's responsibility to advocate meticulous oral care for every patient who develops post-orthodontic WSLs. The meta-analysis for this review implied that both CPP-ACP and fluoride have additional benefits related to routine oral hygiene. The analysis of the current literature also revealed no side effects following the use of these products. However, the findings of this review are still inconclusive because of the inability to include all of the eligible studies in the meta-analysis, the small number of studies and the lack of representation from different settings.
Additional high-quality studies are required to provide more reliable evidence regarding the effectiveness of various remineralising agents and to establish the best clinical practice. Areas to be considered in future trials include standardising the oral hygiene regimen to prevent confounded results. The use of QLF in combination with image analysis would be preferred as both tools allow the measurement of the lesion area. Consequently, crosschecking the results would be made possible to provide more reliable evidence on this topic. 40 As fluoride may have different therapeutic effects for post-orthodontic WSLs at different dosages, including different fluoride dosages in a future study may provide valuable information. 61 Investigating the effect of using CPP-ACP in combination with various forms and dosages of fluoride may also be beneficial. 62 To study the longevity of treatment effects, a longitudinal analysis to measure the difference of treatment results in each follow-up period is recommended.

Conclusion
It may be concluded that simply removing fixed appliances and tooth-brushing with fluoridated toothpaste is effective in reducing post-orthodontic white spot lesions. Furthermore, the current metaanalysis revealed that CPP-ACP and fluoride both improved the enamel mineralisation status (statistically significant). However, the meta-analysis could not accurately establish which remineralising agent, CPP-ACP or fluoride, was more effective due to the heterogeneity in the included studies and with regard to the concentrations and the mode of fluoride delivery. Further high quality randomised controlled trials are required and imperative to establish good clinical practice.