| Sign In to gain access to subscriptions and/or personal tools. |
Improving Clinical Visual Detection—Potential for Caries Clinical TrialsDepartment of Cariology and Endodontics, School of Dentistry, Faculty of Health Science, University of Copenhagen, Denmark; kim{at}odont.ku.dk
ABSTRACT It has been the norm in caries clinical trials to measure caries increment by several different caries-recording systems, including the crude DMF-S/T index. However, there is a reserved attitude as to whether to subdivide the non-cavitated lesions and use arrested lesions in clinical trials. This has been due to the belief that it is not possible to achieve reliable data of the early stages of the disease (Radike, 1972). However, recently, Ekstrand et al.(1997, 1998) showed that it was possible: (1) to differentiate between different stages of non-cavitated occlusal lesions, (2) to differentiate between active and inactive occlusal lesions, and (3) to predict the depth of the lesion. In at least 4 other clinical studies, the reproducibility of recording initial active lesions, cavitated active lesions, and arrested lesions was found to be adequate (Carvalho et al., 1989; Nyvad et al, 1999; Ekstrand et al., 2000; Machiulskiene et al., 2001). Since caries today is a more slowly developing disease in many countries in the world, this will result in prolongation of the duration of the clinical trial, which will increase the costs. As indicated above, there is now sufficient evidence that caries can be clinically diagnosed accurately and reliably in earlier stages as well as in an arrested stage. If such stages of caries are used as outcome variables in caries clinical trials, they may have a positive influence on the trials’ duration and costs.
Key Words: caries • detection • activity assessment • clinical trials INTRODUCTION AND AIMS During the last 30 years, several meetings/workshops have been held concerning requirements for the design, implementation, and analysis of controlled clinical trials within the field of cariology (American Dental Association, 1972; Horowitz et al., 1973; Ainamo, 1982; FDI Commission, 1999). This interest is based on the fact that a well-designed and well-conducted clinical trial with random assignment of participants is a superior method with respect to control for bias, which otherwise can weaken the conclusions (Banting, 1993). However, a proper caries clinical trial is difficult to handle, and since the caries progression rate during the last 30 years has decreased markedly in many countries, it takes longer to verify or reject the stated Ho- hypotheses. This means that the costs of conducting clinical trials are becoming prohibitive, and, according to Angmar-Månsson (2001), this may prevent the clinical testing of potentially effective products. Thus, it is relevant from time to time to examine whether new knowledge about the caries disease in any respect can help us to maintain caries clinical trials as a method to provide evidence of effective products/interventions. One outcome variable in caries clinical trials is caries incidence—the difference between baseline caries prevalence at the start of the trial and the caries prevalence at the intermediate and end of the trial.
Table 1
From the Table, it can be seen that several different criteria have been used to record caries. However, three things are of importance for the purpose of the actual paper: (1) Few trials have subdivided the non-cavitated lesions; (2) few trials have used arrested lesions as an outcome variable; and (3) it is accepted that early lesion development on approximal surfaces cannot be detected clinically (visually), while many clinical trials have incorporated bite-wing radiographs. This paper will provide available evidence concerning the accuracy and reliability of early visible caries signs as indications for caries initiation, progression, and arrestment. The subject of this paper is caries on occlusal and smooth surfaces, exclusively. First, a brief paragraph will describe the patho-anatomical changes that develop during caries development and arrestment. Second, relevant criteria of early caries will be discussed in relation to reliability and accuracy. Third, a description of results obtained in a clinical trial conducted in Moscow (Ekstrand et al., 2000), using a graded caries scoring system including arrested lesions, will be presented. Finally, recommendations will be discussed in relation to how the agent/intervention to be tested in the clinical trial is supposed to interfere with the caries process. PATHO-ANATOMICAL CHANGES IN ENAMEL DURING CARIES INITIATION, PROGRESSION, AND ARRESTMENT Caries is a bacterially induced, slowly developing disease in the dental hard tissues. The disease is caused by cariogenic plaque in which some micro-organisms secrete weak acids when they metabolize carbohydrates. The acids gradually dissolve the underlying mineral, both at the surface and at the subsurface, which causes irreversible structural changes (patho-anatomical changes) in the dental hard tissue. During the 1980s, the working groups around Professor Thylstrup in Copenhagen, Denmark, mapped the patho-anatomical changes in one-, two, three-, and four-week-old caries lesions (Holmen et al., 1985a,b, 1988; Thylstrup et al., 1994). The material consisted of premolars scheduled for extraction causa orthodontica. These premolars were supplied by special orthodontic bands with a standardized space between the bands and the buccal surface to ensure development of dental plaque. The teeth were then extracted 1, 2, 3 or 4 wks after placement of the band. In another setting, the changes in four-week-old lesions were mapped after 1, 2, and 3 wks, respectively, after teeth were debonded, thus with re-exposure of the lesion to the oral environment, including home-based hygiene (Holmen et al., 1987a,b, 1988; Thylstrup et al., 1994). The patho-anatomical changes on the surface enamel were elucidated by scanning electron microscopy, by polarized light examinations, and by macroscopic examinations (clinical changes). The patho-anatomical changes in the subsurface enamel were described by polarized light examinations of ground sections of the lesions.
Caries initiation begins with direct dissolution of the crystals in the surface enamel (Table 2A
Mechanical removal of the cariogenic plaque after debonding leads to surface wear and polishing, which again results in removal of the most porous part of the surface enamel (scanning electron microscopic level) (Table 2B  ). The "new" surface of the arrested lesion (macroscopic level) is therefore hard and shiny (light scatter is less), in contrast to the chalky, rough, and soft surface of the active lesion. The polarized light microscopic examination disclosed that the porosities deep (subsurface) into the arrested lesion were less pronounced than when the lesion was active (Table 2B). Thylstrup et al.(1994) explained this phenomenon with the gradual return to normal pH, promoting reprecipitation of minerals (remineralization) from the internal enamel fluids.
The transition from sound (normal enamel translucency) to opacity visible after or without air-drying can therefore be used as an indication of caries initiation. The transition from opacity after air-drying to opacity without air-drying is an indication of caries progression, and the transition from opacity to less opacity or a maintained degree of opacity is an indication of arrestment of the lesion. The questions, however, are whether these clinical signs and transitions (biologically possible) can be handled reliably in clinical trials, which is a fundamental requirement (Horowitz et al., 1973). To the author’s best knowledge, however, there are no clinical trials which have used such early caries scores (active/arrested). We are therefore forced to seek information about reliability of early caries scores, including arrested lesions, from other types of investigations. EARLY VISIBLE CARIES SIGNS WITH POTENTIAL TO BE INCLUDED IN CARIES CLINICAL TRIALS
Based on the findings by Thylstup and co-workers (see above), Ekstrand et al. (1997, 1998, 2001) and Ricketts et al.(2002) created and evaluated a scoring system to detect occlusal lesions, predict their depth, assess their activity, and, finally, predict whether the lesion is infected. Table 3
The in vivo studies (Ekstrand et al., 1998; Ricketts et al., 2002) used third molars scheduled for extraction, so it was histologically possible to validate the activity and the level of infection of the individual lesions. Just before extraction, the teeth were visually examined according to the scoring system in Table 3  (column 1). Scores 1 and 2 were subdivided, however, into scores 1 and 1a and 2 and 2a (column 3), where white spots indicated activity of the lesions (Holmen et al., 1987a), and brown-spot lesions were supposed to indicate lesions in an arrested stage (Thylstrup and Fejerskov, 1994). As validation criteria, the dye methyl red was used (column 4). This dye was applied to the section faces of ground sections through the lesions. Methyl red is a pH-indicator and changes color at a pH of 5.5 (Holtzclaw et al., 1991), which is the level where the hydroxyapatite dissolves (Larsen, 1975; Larsen and Bruun, 1986). When the visual scores 1, 2, 3, and 4 were used as indicators for active lesions, and scores 0, 1a, and 2a as indicators for sound or inactive lesions, a strong and significant association was observed with the histological validation (C = 0.88, p < 0.001; Ekstrand et al., 1998). This pattern was confirmed when the ground sections were examined in polarized light and, based on this appearance, were rated subjectively as active or inactive according to the system proposed by Kidd (1983). In accordance with Thylstrup and Qvist (1986), lesions with moderately or heavily infected dentin were cavitated (Table 3, column 5). Nyvad et al.(1999) developed a caries diagnostic system (differentiating early lesions from cavitated lesions and active lesions from arrested lesions) which could reliably differentiate among 10 scores in a suboptimal clinical setting in Kaunas, Lithuania. Thus, Kappa values ranged between 0.74 and 0.85 for intra-examiner examinations, and between 0.78 and 0.80 for inter-examiner examinations. The biggest problem was consistency in the differentiation of sound from non-cavitated lesions (active as well as inactive). Recently, the same recording system has been used in a three-year clinical trial (Machiulskiene et al., 2001), and the intra-examiner reliability, which was tested each year, ranged from substantial to excellent. Since both studies were clinical, it was initially not possible to test the accuracy of the scoring system in terms of predicting depth and activity of the individual lesions. However, by using construct validity (the extent to which the measurement corresponded to theoretical concepts concerning the phenomenon under study), the authors concluded, in a more recent study (Nyvad et al., 2003; Nyvad, 2004), that the diagnostic criteria were valid for assessing caries activity. It should be underlined, however, that the system was not a strict visual classification system, since the probe was used for characterizing the surface as smooth, rough, or with breakdowns. Thus, there is evidence that early caries signs can visually be recorded in a reliable way, and severity and activity of the lesion can be predicted accurately. However, optimal clinical conditions are required, including cleaned teeth, compressed air for drying the teeth, and a dental light. EXPERIENCE WITH A CLINICAL TRIAL WHERE NON-CAVITATED ACTIVE AND ARRESTED LESIONS WERE RECORDED AS OUTCOME VARIABLES ALONG WITH THE NORMAL DMF-S INDEX
As seen in Table 1 With this recording system, it is possible to express both the traditional DMF-S/T values and the number of non-cavitated active lesions (D1), arrested lesions, and sealed surfaces. In fact, the number of teeth which were probably in need of root treatment could be expressed. When the clinical trial in Moscow was initiated, the six-year-olds in both the study and the control groups (N = 50 in each group) already had a number of D1 lesions (mean D1-S, study group = 2.34 vs. 1.42 in the control group), but very few lesions at a cavitated level or restored surfaces (mean DMF-S, study group = 0.06, 0.19 in the control group). At the end of the trial, the control group had a mean DMF-S = 2.24, but a mean D1-S = 5.92. The actual caries experience increased in this case by a factor of 4. In the study group, the mean DMF-S was finally 0.28, but the mean D1-S was 1.66. Thus, the mean DMF-S increased slightly during the study period, while the number of D1 lesions dropped, which corresponded to an increase in the number of arrested lesions that, on average, was 2.66 surfaces at the end of the clinical trial. As suggested by Nyvad et al.(1999), evaluation of the effects of various non-operative/preventive interventions of caries requires a diagnostic system which reflects the dynamic nature of caries at all stages of lesion progression. The recording system used in the clinical trial in Moscow fulfilled these requirements. DISCUSSION
At the international workshop ("Understanding dental caries—impact for future oral health care") held in Aarhus, Denmark, in 1996, Kingman and Selwitz (1997) suggested what they called a "new paradigm" for assessing caries initiation and progression. They proposed to incorporate two levels of caries detection: incipient lesions, denoted D1, and frank caries, denoted D2. They also suggested including two levels concerning fillings: small (F1) and large (F2). Since the reliability of detecting various stages of early lesions is adequate under optimal conditions (Ekstrand et al., 1995, 1997), this review argues that incipient lesions (D1) can be further subdivided (D1,1, D1,2). Systems differentiating between opacity visible only after prolonged air-drying and opacity visible on wet surfaces (Ekstrand et al., 1997), or, as suggested by Marthaler (1965), first detectable changes and distinct changes without loss of surface continuity (Table 1 This review also suggests the use of arrested lesions and biologically possible reversals as, e.g., transitions from opacity visible on a wet surface (D1,2) to opacity visible only after air-drying (D1,1), or from these signs to sound, as special outcome variables. Also, lesions which are stable during two clinical assessments could be used as an outcome variable. Inclusion of arrested lesions and biologically possible reversals are of particular importance when clinical trials are planned to be conducted in countries with low caries progression rates, and if the agents/interventions to be tested influence the etiology of caries, as, e.g., does plaque (various anti-bacterial products).
FOOTNOTES Presented at the International Consensus Workshop on Caries Clinical Trials, Glasgow, Scotland, January 7–10, 2002 REFERENCES
Journal of Dental Research, Vol. 83, No. suppl 1,
C67-C71 (2004) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
> 0.72) (
years of artificial fluoridation of drinking water in the Netherlands—The Tiel-Culemborg experiment. Arch Oral Biol 5:284–300.
