Currently, dental practitioners are using air abrasion as an aid in diagnosing suspected carious lesions.³ When used in this manner, many authors recommend eliminating darkened pits and fissures.^4,5 Removing stain and organic debris allows better visualization of the depth of the pits and fissures and aids in detecting decay. This pit-and-fissure cleaning invariably removes a small amount of tooth structure, which should be restored with a sealant if it is within enamel or flowable composite if it penetrates into dentin.

It is well-known that sealants and other composite materials require periodic maintenance or replacement.^6,7 This need for continuing treatment leads to the question of whether a patient’s oral health is improved by early preparation of questionable carious lesions. When to prepare an incipient lesion is not a new question.⁸ With the proliferation of air-abrasion units, there is a definite trend toward treating or at least performing an "enamel biopsy" of these questionable lesions earlier than in the past. The justification is that occlusal caries has become more difficult to diagnose, which some have suggested is the result of an increased use of tooth-paste containing fluoride.^9,10 This difficulty of diagnosis has spawned the term "hidden caries" to describe occlusal carious lesions that are detectable through radiography but not through routine clinical examination. Yet there has been, overall, a reduction in the rate and progression of caries owing to an increased use of fluoride.¹¹

One author has suggested treating caries similar to cancer³—that is, treat it early while it is small and conserve tooth structure with new microabrasion techniques. Others disagree, noting that the greater availability of fluoride in various forms reduces caries and can enhance remineralization.¹² Given these competing views, we proposed a five-year randomized clinical trial to evaluate the risks vs. the benefits of early treatment of questionable incipient carious lesions using air-abrasion technology. The study will evaluate two null hypotheses:

– There is no difference in the number of carious lesions in dentin diagnosed and treated in the control group (which was observed but not treated) vs. the experimental group (which received early treatment with air abrasion).

– There is no difference in the volume of tooth structure removed when placing restorations to restore carious lesions that penetrate into dentin in the control group (observed) vs. the experimental group (early treatment with air abrasion). We felt that by treating questionable carious lesions early instead of waiting until they were clearly carious, the size and the number of cavity preparations into dentin would be reduced.

	METHODS AND MATERIALS

TOP ABSTRACT METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Patient selection. We used a patient assent form for minors and a patient consent form for adults and parents or guardians, both of which were approved by the institutional review board of the University of Michigan, Ann Arbor, before we recruited any patients.

We enrolled 93 patients aged from 12 to 36 years who had among them 223 questionable carious lesions; all patients were drawn from the general dentistry clinics at the University of Michigan School of Dentistry. These patients were selected because, during a routine dental examination, they were diagnosed with at least one questionable carious lesion in the pits and fissures of a posterior tooth. The teeth selected were free of frank caries (softness at the base of a pit or fissure, decalcification or cavitation) or evidence of radiographic caries. Most of the teeth selected had deep staining (Figure 1) or explorer retention in a pit or fissure. We included no more than three teeth from any given patient to preclude major losses to follow-up if patients dropped out of the study for any reason and to reduce patient bias in caries risk.

View larger version (81K): [in this window] [in a new window]   Figure 1. One of 223 posterior teeth enrolled in the controlled clinical trial with stained pits and fissures.

Of the 113 teeth with questionable incipient carious lesions that were air-abraded and restored in the treatment group, 50 (44 percent) had caries extending into dentin.

Baseline evaluation. After enrolling in the study, each patient was examined independently by two dentists (any two of the three dentist authors). Each dentist used magnification (x2.5) to evaluate each study tooth for gingival health,¹³ plaque retention,¹⁴ caries, darkest color found in the pits and fissures, and explorer retention in the pits and fissures with the same explorer. One examiner took an impression of the occlusal surface of each enrolled tooth with a clear polyvinyl siloxane, or PVS, bite registration material and charted all decayed, missing and filled teeth. After the examination, if there were any disagreements between the two examiners, consensus was reached through discussion. After all evaluations were completed, the research coordinator used a table of random numbers to assign each enrolled tooth independently to either a treatment group or a control group.

Treatment group. We isolated the teeth randomized into the treatment group with a rubber dam and abraded the questionable pit-and-fissure system using a dental abrasion system and aluminum oxide powder with a mean particle size of 27 micrometers at 80 pounds per square inch (5.6 kilograms per square centimeter) to remove stain and any associated carious tooth structure. The patients were told that most patients do not require local anesthetic for air abrasion, but they were encouraged to ask for it at any time. To patients who requested anesthetic, we administered 2 percent lidocaine with 1:100,000 epinephrine using standard dental techniques.

We took care not to overprepare the teeth by stopping repeatedly to assess visually and by touch the degree of stain and caries removal. We considered using caries detector dyes, but decided against it owing to the many conflicting studies concerning their efficacy.^13–15 After removing all stain and any associated caries, we made an assessment as to whether the preparation extended into dentin, an indication that the tooth was in need of operative intervention.

If the preparation did extend into dentin, we made an impression to assess its volume with a fast-setting low-viscosity PVS material. We used the impression of the occlusal surface made at baseline to form the occlusal surface of the preparation impression (Figure 2). A similar impression was made of any preparation in a control tooth that extended into dentin subsequent to being diagnosed and treated with air abrasion for occlusal caries. Each year, all impressions will be weighed by one technician, who will be blinded as to which impressions originated from the treatment or control teeth. These impression weights, a surrogate measure for the volume of tooth structure lost, will be compared using a mixed-model analysis of variance, or ANOVA. This will indicate if there is a significant difference (P < .05) in size of the preparations prepared in teeth with questionable carious lesions and those prepared in teeth with diagnosed carious lesions.

View larger version (117K): [in this window] [in a new window]   Figure 2. A preparation impression (green), not trimmed of flash, shown adjacent to the clear impression of the occlusal surface taken at baseline.

The preparations were restored using the following procedures. We placed glass ionomer lining cement if the preparation extended into the inner one-third of dentin, since it has been shown to reduce the chance of secondary decay in vitro.^16,17 This step was necessary in only three restorations. We etched the preparation and the tooth surface 1 millimeter beyond the cavosurface margin with 37 percent phosphoric acid gel, rinsed it, dried it and applied and cured a dentin-enamel bonding agent according to the manufacturer’s instructions. We placed a flowable light-cured composite (Tetric Flow, Ivoclar Vivadent) using a unit-of-use tip and taking care not to trap air in the narrow preparations (Figure 3). After light-curing, we removed the rubber dam and checked and, if necessary, adjusted the occlusion. Two examiners (any two of the three dentist authors) used a modified set of Ryge criteria¹⁸ (Table 1) to evaluate the finished restoration (Figure 4) independently for color, margin discoloration, margin adaptation, anatomical form and surface texture. If there was any disagreement between examiners, a consensus was reached after discussion.

View larger version (90K): [in this window] [in a new window]   Figure 3. An air-abraded preparation in one of 113 teeth randomized into the treatment group at baseline.

View larger version (99K): [in this window] [in a new window]   Figure 4. Completed restoration at baseline.

Statistical evaluation. The baseline demographic and examination data for the treatment group was analyzed using logistic regression with generalized estimating equations for clustered data to determine which baseline factors were associated with caries extending into dentin. Before the end of the five-year clinical study, we will compare the number of teeth discovered with caries penetrating into dentin in the control and treatment groups using the ² test, and we will determine the probability of a control tooth’s becoming carious using the Kaplan-Meier method. Using a mixed-model ANOVA, we will compare the treatment and control preparation impression weights in a surrogate measure of volume to determine if tooth structure was conserved by early treatment.

	RESULTS

TOP ABSTRACT METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Of the 113 teeth with questionable incipient carious lesions that were air-abraded and restored in the treatment group, 50 (44 percent) had caries extending into dentin. After 12 months of clinical service, we found three preventive resin restorations that exhibited a partial loss of sealant but did not require any re-treatment. Two preventive resin restorations with penetrating staining were re-treated.

In the control group for teeth evaluated at 12 months, the probability of caries was 11 percent (95 percent confidence interval, 4 percent to 18 percent). Using the ² statistic to compare the number of teeth diagnosed with caries extending into dentin in the treatment and control groups, we found a very significant difference between the groups (P < .001). This indicates that after one year, the control group had significantly fewer carious lesions diagnosed than were determined by operating on the treatment group. There was no statistically significant difference (P = .279) between the weight of the treatment preparation impression (0.027 gram) and control preparation impressions (0.020 g). Table 2 presents the distribution of tooth type in the treatment and control groups, along with the mean ages of the patients in both groups whose teeth had caries extending into dentin.

– sex;

– age;

– fluoride history;

– tooth type;

– decayed, missing or filled surfaces;

– score on the Löe and Silness Gingival Health Index¹⁹;

– score on the Simplified Oral Hygiene Index²⁰;

– pit and fissure color;

– explorer retention of the pits and fissures.

Caries penetrating into dentin was positively correlated with explorer retention (P = .006)—that is, the more the explorer was retained in a pit or fissure, the more likely it was that caries had extended into dentin. Age was negatively correlated with caries extending into dentin (P = .0313); the greater the age of the patient, the less likely it was that caries had progressed into dentin.

In the cases of 13 (12 percent) of the 113 teeth in the treatment group, the patient requested local anesthetic during tooth preparation. In the case of one (11 percent) of the nine control teeth diagnosed with caries at 12 months and treated with air abrasion, the patient requested local anesthetic. All but one of these teeth were associated with deeper preparations.

	DISCUSSION

TOP ABSTRACT METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSION REFERENCES

 
By providing early treatment for 113 questionable carious lesions in the treatment group, we discovered 50 carious lesions (44 percent) that extended into dentin. Eight of those 50 (16 percent) had more than minimal extension into dentin. Surprisingly, none of the control teeth diagnosed with caries at one year had more than minimal caries extending into dentin. This likely affected the volume comparisons of the preparations in the treatment and control teeth, and led to a larger mean preparation size for the treatment teeth. It would be expected that the number and size of restoration preparations in the control group would increase in the future, but recalls we have carried out since the 12-month point do not indicate any increase in the mean size of preparations.

The enrollment of subjects from the patient pool receiving dental services at the University of Michigan School of Dentistry may be subject to selection bias. Patients receiving or requesting treatment may be more sensitive to oral health care issues than the general population, which could bias the results toward a reduced caries risk. On the other hand, patients attending the university clinics pay reduced fees and may represent a lower socioeconomic status than those typically visiting a private practice, which could bias the results toward an increased caries risk. It is unclear how these factors affect the generalizability of the results. Since the ages of the subjects were from 12 to 36 years of age, with a mean age of 23 years at baseline, we feel that this group of patients is younger than those in the average private general practice.

Before the start of the study, and after discussions with cariologists, we believed that approximately 25 percent of the teeth with questionable incipient carious lesions in the control group would become carious each year. After completing baseline procedures and finding that 44 percent of the teeth in the treatment group had caries that had progressed into dentin, we felt that more than 25 percent of the control teeth would become carious before their one-year recall visit. There should be approximately an equal percentage of control teeth with caries extending into dentin owing to the randomization of teeth at baseline. Knowing this may lead to a bias to diagnose caries before it has progressed into dentin (a false-positive). This did not occur during the first year of recall, as all control teeth diagnosed with caries had decay extending into dentin when they were air-abraded.

The idea that the evaluators are not diagnosing caries when caries is present (false-negative) is supported by two facts. First, none of the control-group patients diagnosed with a carious control tooth at recall had any sensitivity before the recall appointment. Second, the percentage of patients who requested local anesthetic was similar in the control group (11 percent) and the treatment group (12 percent).

Since each tooth at baseline and all control teeth at each recall visit are examined with an explorer independently by two examiners to assess explorer retention of the pits and fissures, the teeth in this study are probed more often than they would be in a private practice setting. This could allow the intrasubject transfer of cariogenic bacteria from one control tooth to another and produce a higher rate of caries than expected. Also, the act of probing the occlusal surface with an explorer has been shown to cause irreversible traumatic defects in demineralized areas and favorable conditions for carious lesion progression.²¹ This mechanical damage leads to an increased demineralization of the enamel.²² Therefore, if this effect occurred, it would be expected that the control group would have a higher caries rate than expected. This did not happen at the 12-month recall point.

Bayne and colleagues²³ reported that the mechanical durability of flowable resin-based composite was 60 to 90 percent of that of conventional resin-based composites. This could raise a concern about the longevity of the restorations placed in this study. At this early stage in the study, there is no sign of occlusal wear. The narrow preparations could limit the direct contact between these restorations and the opposing cusps.

	CONCLUSION

TOP ABSTRACT METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSION REFERENCES

 
In this randomized clinical trial, we compared the size and number of air-abraded preparations that extended into dentin in an early treatment group and a control group. Judged on the basis of the weight of the preparation impression (a surrogate measure of volume), there was no benefit gained from early treatment of questionable carious lesions. Whether longer periods of service will show a different result remains to be seen. Because many more teeth were air-abraded in the early treatment group, and because those restorations are associated with additional costs in terms of money, time and future maintenance, we believe that early operative intervention should not be recommended until proven clinical benefits have been demonstrated.

View larger version (149K): [in this window] [in a new window]   Dr. Hamilton is an assistant professor, Department of Cariology, Restorative Sciences and Endodontics, University of Michigan School of Dentistry, 1011 N. University, Ann Arbor, Mich. 48109, e-mail "jchamilt{at}umich.edu". Address reprint requests to Dr Hamilton.

View larger version (128K): [in this window] [in a new window]   Dr. Stoffers is a clinical assistant professor of dentistry, School of Dentistry, University of Michigan, Ann Arbor.

View larger version (142K): [in this window] [in a new window]   Dr. Dennison is a professor of dentistry, School of Dentistry, University of Michigan, Ann Arbor.