Brian H. Clarkson, Ph.D., M.S., L.D.S., and Mary Rafter, D.D.S., M.S.:
The purpose of our review was to appraise and synthesize the relevant literature on several questions pertaining to the prevention of dental caries:
1. This question was broken down into four subquestions by treating enamel, dentin, increasing remineralization, and increasing resistance to demineralization as separate entities. A further breakdown was conducted under the headings human (clinical), animal, and laboratory studies.
2. Is lased enamel or dentin more, or less, susceptible to demineralization, compared to nonlased enamel and dentin?
For this question only laboratory studies were found, and enamel and dentin were treated as separate questions.
3. Do fluoride-releasing dental materials increase the remineralization of demineralized (carious) human enamel or dentin, or increase the resistance to demineralization (caries) of these tissues?
Only human clinical trials and human in situ studies were reviewed in addressing this question. Enamel and dentin were treated as different subjects of inquiry, as were remineralization and demineralization. Only studies reporting direct measures of changes in enamel and dentin remineralization and increased resistance to demineralization were considered in the appraisal. Investigators using such indirect measures as, for example, fluoride uptake or plaque accumulation were excluded.
A further question more closely linked to repairing dentinal caries and not as relevant to caries prevention was also considered:
4. Since no human studies have been reported and in vitro studies did not show tubular dentin formation, only animal studies were reviewed.
Methods
A search was made of articles published in peer-reviewed journals, written in English, and indexed in MEDLINE or EMBASE. References in review articles were also used as a source if they were not cross-referenced in MEDLINE or EMBASE. All articles had to be published after 1976. The databases were searched using appropriate key words for each of the questions asked. Use was also made of the caries hedge setup for reviewers involved in literature review for this Consensus Development Conference. References in the reviewed articles were also searched for other relevant reports.
The two investigators independently read all the abstracts from the MEDLINE, EMBASE, and hand searches. Relevant reports were then tagged. Discrepancies between the investigators were resolved by consensus after a further reading of the disputed abstracts. Articles with tagged abstracts were then photocopied and distributed equally between the two investigators, except for articles on bone morphogenic protein, which were read by only one investigator (Clarkson).
All articles were then abstracted and entered into the evidence tables under various headings, and then scored, except for those on studies that had no controls. Purely descriptive studies on BMP activity were included, but descriptive studies in which statistical analysis was deemed appropriate but was not carried out, and articles in which conclusions were drawn from inappropriate statistics, were not.
The scoring system was an all-or-none system based on the evidence table headings. If information was available in the article under the heading, it was given a score of 1; if it was missing, it was given a score of 0. Publication date, author�s name, and study type were not included in this scoring system, nor was the information under the heading "findings." The total score differed for each question. The score assigned, and the possible total score for each article, are given in the last column of the evidence table. All articles were scored independently, and disagreements were resolved by consensus.
Results
For question 1 (partitioned toothpastes), only 12 of the 35 abstracts dealt with a toothpaste in which the calcium was separate from the phosphate and fluoride until ions were delivered to the tooth surface. Of these, seven were in vitro investigations, three were animal studies, and two were clinical trials. After the full articles were read, one in vitro study and one clinical study were excluded because of insufficient data.
All of the studies that were included had one author�s name in common, but the research was carried out at several different institutions. One of the animal studies and six of the in vitro studies dealt with remineralization of enamel, but none of the studies reported on remineralization of dentin. The other two animal studies tested the partitioned toothpaste�s ability to increase the resistance of enamel to demineralization. There were no studies of dentin resistance to demineralization. The one clinical trial tested the partitioned toothpaste�s ability to inhibit both coronal and root caries. All these studies showed positive results except for the clinical trial, in which the partitioned toothpaste reduced root caries but not coronal caries. Thus, in all but the class I (as designated by AHRQ�s U.S. preventive service task force grading of the evidence) clinical trial investigating enamel caries, the partitioned toothpaste showed either greater remineralizing enhancement or greater increase in resistance to demineralization of enamel and dentin compared to a nonpartitioned toothpaste containing calcium, phosphate, and fluoride.
Question 2 focused on the demineralization potential of lased versus nonlased enamel or dentin. Of the 84 abstracts initially read, 14 in vitro studies were evaluated. Seven of these were excluded because of no, or inappropriate, statistics. Of the seven remaining, five out of six concluded that lased enamel was less soluble than nonlased, while the one article on lased dentin reached the same conclusion.
Question 3 asked whether fluoride-releasing restorative materials increase the remineralization or the resistance to demineralization of enamel or dentin. Of the eight clinical trials, two were excluded because there were no control groups. Of the remaining six, one was designated a class 11-1 study, four were class 11-2 studies, and one was a class 11-3 study as designated by the U.S. preventive services task force grading. Of the six in situ studies that were also reviewed, two were excluded, one because it used bovine tissue and one for incomplete data. All but one of the clinical and in situ studies were short-term�that is, less than 16.3 months. The other lasted 3 years. They used a variety of methods for measuring remineralization and resistance to demineralization of both enamel and dentin. The study participants (or specimens) were also subjected to several different caries challenges. Eight of the 10 studies did not report on examiner calibration or reliability. Of the six clinical trials, five dealt with enhancing the resistance of enamel to demineralization and one dealt with dentin remineralization. No clinical trials on enamel remineralization or increasing dentin resistance to demineralization are discussed here, either because no studies had been conducted or those that had been conducted did not meet our criteria. Of the four in situ studies, two dealt with increasing the resistance of enamel to demineralization, one looked at both increasing the resistance to enamel demineralization and enhancing enamel remineralization, and one looked at dentin remineralization. No in situ studies on enhancing the resistance of dentin to demineralization were found.
In the five clinical trials investigating the effects of fluoride-releasing materials in enhancing enamel�s resistance to demineralization, four recorded increased resistance and one showed no difference between the experimental and control groups. In the one study on increasing the remineralization of dentinal lesions with these materials, no difference was seen between the experimental and control groups. The one study that looked at remineralization of enamel in conjunction with increasing enamel resistance to demineralization failed to state the remineralization results. Of the in situ studies, the one study investigating the remineralization of dentin by fluoride-releasing materials showed increased remineralization, while the three examining enamel resistance to demineralization all recorded increased resistance.
For Question 4 on BMP�s ability to stimulate dentin formation, all six articles reviewed were animal studies. Irrespective of the species, all showed that BMP stimulated new dentin formation. The reparative dentin included both tubular and nontubular (osteo) dentin. One study that tested transdental transition of BMP showed that BMP activity did, in fact, cross dentin. Two of these studies used a crude BMP extract, while four used BMP-7 (OP-1).
Conclusions
Question 1:In spite of the fact that all the studies on using the partitioning of the active ingredients of toothpaste had one author in common and that only a few studies had been conducted, there is sufficient evidence from the animal and in vitro studies to suggest that this technology has promise in enamel caries prevention. In humans, however, the sole class I clinical trial did not show a difference in enamel caries reductions between experimental and control groups in a high-risk population. But in the same study the partitioned toothpaste prevented root caries to a greater extent than a conventional toothpaste. Independent, randomized, controlled clinical trials need to be conducted to determine if this therapy�s usefulness can be generalized to all population groups. Studies also need to be conducted on its usefulness for preventing dentin caries.
Question 2: In vitro testing of the solubility of lased enamel has demonstrated that it is less susceptible to demineralization than nonlased enamel. The results for dentin were similar, but only two studies met the criteria for inclusion in this review. Further in vitro investigations to determine if lased dentin is indeed less soluble should be undertaken.
The reviewed studies used several different laser types, application times, laser wavelengths, power, demineralization models, and target distances (i.e., distance from laser head to tissue) and made it impossible to recommend a standard procedure. Investigations should be performed to establish the standard protocol for application in clinical trials that must be completed before this therapy can be recommended for caries prevention.
Question 3: The small number of studies using direct measures of caries prevention and the short duration of those studies made it impossible to draw any conclusions about the long-term benefits of these measures. Randomized, controlled clinical trials need to be conducted over a period of at least 2 years to answer the four subquestions reviewed in this paper�whether fluoride-releasing dental materials increase the remineralization of carious enamel and dentin, and whether these materials increase the resistance of enamel and dentin to caries.
Question 4: All of the animal studies reviewed reported that crude BMP extracts and BMP-7 were able to regenerate dentin (tubular and atubular) when placed on vital pulps. One study also showed that the active signaling molecule can cross dentin and stimulate a pulpal response. One anecdotal report of a clinical trial using BMP-7 suggested that the results of the study were equivocal. Nevertheless, the animal studies suggest that this therapy provides positive results. Further investigations should be undertaken, controlling for the drug carrier and studying the effect of inflammation on the BMP-7 activity. After these animal studies are completed, human clinical trials should be conducted.