For WSL diagnosis, visual-tactile examination is used in clinical practice. Of enormous importance, in addition to early detection of lesions, is the clinical practicality, reproducibility, and diagnostic value of a method. In the present clinical study, the results of visual-tactile examination were compared with those of QLF measurement. Both diagnostic methods independently documented an increase in the prevalence and incidence of WSL within one year following the commencement of orthodontic treatment (Figs. 1 and 2). After the insertion of the MB, the S. mutans population increases sharply after a period of three months (19), as the apparatus promotes adherence and thus colonization of the bacteria (20). Consequently, the observation period of our study corresponded to a duration that gives caries-promoting factors, which are elevated during the active therapy phase (10), sufficient time for the initial lesions to form.
The prevalence of WSL during orthodontic therapy has been investigated in several studies and varies widely in the results. The large variation between 2% − 96% (1, 9, 11–13, 21) can be explained by the fact that there were significant differences in study design and analytical approach (22, 23). The studies differ not only in treatment duration and age of the subjects, but also in the type of diagnosis.
In some studies, intraoral photographs were used to apply the index and to detect the WSL (14, 21). In other studies, the index was determined directly by visual inspection of the subject (9, 11, 24). Furthermore, ethnicity, subject selection and the use of different materials with regard to bonding and oral hygiene are known to influence the development of WSL (22, 25). These methodological differences may have an impact on the study results. However, the studies are congruent with regard to the result that there is an increase in WSL under orthodontic treatment.
The distribution pattern of the lesions in relation to the individual teeth revealed the highest prevalence in the anterior maxillary region. According to our study, in particular the central incisors are most frequently affected by WSL (Figs. 1 and 2). The reason for this may be the early eruption of the incisors, which are thus exposed the longest to the cariogenic factors of the oral cavity (26). The lowest prevalence values were detected for the second premolars in the maxillary and mandibular regions. In addition to the anatomical proximity to the excretory duct of the parotid gland and the associated buffer function of saliva as an essential component of caries protection, another reason for the low prevalence values may be the assessability of the teeth by the QLF method, which decreases from anterior to posterior. Hence, the selection of teeth to be imaged by the biluminator should be critically considered. The continuous assessability of the anterior teeth is much higher than that of the premolars. The buccal surface of these teeth was most frequently recorded as "not assessable" in our study. Poor image quality, an only partially visible surface due to gingival hyperplasia or a not yet fully erupted tooth led to the exclusion of some tooth surfaces.
Comparing the diagnostic methods, no lesion was detected by either QLF or EDI on 970 of 1374 tooth surfaces examined (70.6%). On 259 surfaces (18.9 %), a lesion was detected by both methods (Fig. 3). The agreement of the measurement results of both methods with respect to the detection of WSL and fluorescence loss was 89.1% (n = 1229) with a Cohen’s kappa coefficient of 0.71, which, according to Fleiss, indicates good agreement (27) and one can assume a strong correlation of the two methods. (28). One advantage of the QLF method is that the dark appearance of the lesions makes them visible to the subjects and thus encourages them to practice more effective oral hygiene (17). Consequently, the subjects' motivation and cooperation with regard to brushing behavior increases, as they have the opportunity to follow the development and appearance of their lesions on the monitor (29). In our study, a lesion was detected on 79 tooth surfaces (5.8%) by QLF only. EDI findings revealed negative WSL on these surfaces. Heinrich-Weltzien et al. (28) assume that due to the more sensitive QLF diagnostics, a WSL on smooth surfaces can be detected at an earlier stage. According to the present study, the average QLF values of these lesions are higher than the values of lesions already detected by the naked eye. They correlate with the results of a comparative study (EDI + QLF = ΔF -9.6%; and QLF = ΔF -7.1%) (28).
The clinical relevance and resulting consequence of such findings should be critically considered. They are not comparable to study results relating QLF data to International Caries Detection and Assesment System Score (ICDAS) and a histologic score (30). The aim of this study was to provide a feasible framework for the validation of QLF scores (31). An ICDAS-II score of 0 (no visible caries after air drying) corresponds to a QLF score of 0 (ΔF = -0.5 to -12) and this in turn corresponds to healthy tooth structure. An ICDAS-II score of 1 (first visual changes in the enamel surface, which are only visible after the tooth has dried and may appear in the form of opacities, whitish or brownish discolorations) corresponds to a QLF score of 1 (ΔF = -12.5 to -18) according to Alammari et al. (30). Furthermore, the QLF values were supported by histological examinations. At values between − 10.5 and − 15, initial visual enamel changes may be visible. From a value of -15.5, clear visible changes can be seen (31).
The correct drying of the surfaces is a very important part of both examinations. The QLF recordings take longer than normal recordings due to the fact that first photographs and then the recordings are taken. A complete absence of saliva on the tooth surfaces to be recorded is essential for an adequate QLF analysis, but proved to be difficult to perform and very time-consuming in some subject cases due to the coordination of the tongue, cheek retractors and the correct edge to edge positioning of the teeth. Increasing dehydration of the enamel also affects the clinical QLF analysis, as this results in a greater fluorescence loss (32).
Heinrich-Weltzien et al. (28) stated in their study that 4.9% of the detected lesions were only detectable by visual findings but not by the QLF method. The reasons given were the known confounding factors of QLF analysis, such as gingival hyperplasia due to inflammation, a tooth surface covered with plaque, or poor image quality (focus/overexposure). In our study, these factors were ruled out through professional dental cleaning and the exclusion of tooth surfaces that could not be adequately assessed by either method; however, according to Heinrich-Weltzien, the detection of lesions at the gingival margin using QLF also proves to be difficult. If a lesion was suspected by the appearance of a darker area on a tooth surface, but parts of it were obscured by the arch wire, it could not be measured correctly by the program. However, in-situ diagnosis using EDI is possible despite the arch wire by using mirrors and changing the viewing angle. In such cases, the QLF method had limitations. This may be a reason for the fact that on 4.8% (n = 66) of the surfaces a lesion was detected by way of visual examination only (using EDI), but no fluorescence loss could be measured on this area by QLF because this area was obscured by an arc.
The acquisition of QLF images must be performed by trained personnel, since identical angulation of the images is necessary with regard to the monitoring of lesions in order to ensure correct progression analysis by the software as well as the quality and diagnostic value of the images. In order to be able to use QLF reliably in everyday clinical practice, it is essential to minimize the interfering factors that have a great influence on the diagnostic value of this procedure.