Journal of Dental Research

In the past 10 to 12 years, there have been several statistical issues identified in periodontal research which require and have generated non-standard or new statistical approaches. The purpose of this paper is to give an overview of these issues and approaches. Three general categories of issues are described: (i) statistical methods for detecting when disease progression occurs, and biological theories and corresponding statistical models which attempt to describe how the disease progresses; (ii) design issues in studies of therapeutic efficacy; and (iii) analytic issues arising from periodontal data analysis.

The purpose of this study was to assess the prevalence and pattern of self-reported TMD jaw pain in a randomized stratified sample from the general population living in the Province of Quebec, Canada. Through a telephone survey, standardized questions were asked to 897 French-speaking respondents, aged 18 years old and over, regarding frequency, severity, daily pattern of jaw pain, presence of difficulty in opening, joint clicking, and sleeping problems. All prevalence estimates were adjusted to the sociodemographic distribution of the non-institutionalized population. The results indicate that TMD jaw pain is self-reported by 30% of the general population; however, the prevalence of cases reporting frequent episodes (quite often or very often) is estimated at 7%, with more than two-thirds (69%) of the respondents in this subgroup experiencing moderate to severe pain. The prevalence rates of frequent difficulty in opening and joint clicking were estimated at 9% and 4%, respectively. Approximately one in four subjects with frequent episodes of jaw pain also reported frequent joint clicking or difficulty in opening, and a strong association (Gamma coefficient > 0.6) was found among all three TMD symptoms. Our data suggest that the prevalence of clinically significant TMDrelated jaw pain (frequent jaw pain of moderate to severe intensity) is approximately 5% in the general population of the Province of Quebec. In the nine months preceding the survey, about 2% of the total population sought treatment for a TMD symptom.

As a rule, cast restorations do not allow for free curing contraction of the resin composite luting cement. In a rigid situation, the resulting contraction stress is inversely proportional to the resin layer thickness. Adhesive technology has demonstrated, however, that thin joints may be considerably stronger than thicker ones. To investigate the effects of layer thickness and contraction stress on the tensile strength of resin composite joints, we cured cylindrical samples of a chemically initiated resin composite (Clearfil F2) in restrained conditions and subsequently loaded them in tension. The samples had a diameter of 5.35 mm and thicknesses of 50, 100, 200, 300, 400, 500, 600, and 700 μm, 1.4 mm, or 2.7 mm. None of the samples fractured due to contraction stress prior to tensile loading. Tensile strength decreased gradually from 62 ± 2 MPa for the 50-μm layer to 31 ± 4 MPa for the 2.7-mm layer. The failures were exclusively cohesive in resin for layers between 50 and 400 um thick. Between 500 and 700 um, the failures were cohesive or mixed adhesive/cohesive, while the 1.4- and 2.7-mm layers always failed in a mixed adhesive/cohesive mode. For the resin composite tested, the contraction stress did not endanger the cohesive strength. It was concluded that if adhesion to tooth structure were improved, thinner adhesive joints might enhance the clinical success of luted restorations.

This study aimed to determine the degree of eccentricity between different tungsten carbide bur manufacturing techniques and to study the effect of bur inaccuracy on dental enamel. Error in bur concentricity may arise from malalignment of the steel shaft and carbide head in a two-piece construction bur. Cutting blades rotate at multiple radii from the shaft axis, potentially producing vibrations and damage to the cut substrate. Techniques now allow for the manufacture of one-piece tungsten carbide burs with strength adequate to withstand lateral loading. A comparison of tungsten carbide dental cutting tools revealed the true extent of concentricity errors. Variation in alignment of the cutting head and shaft in the two-part constructions incurred between 20 and 50 μm of additional axial error. High-speed cutting interactions with dental enamel between carbide burs were studied by means of a video-rate confocal microscope. A cutting stage fitted to a Tandem Scanning Microscope (TSM) allowed for real-time dynamic image acquisition. Images were captured and retrieved by means of a low-light-level camera recording directly to S-VHS videotape. Videotape showing the interactions of high-speed rotary cutting instruments (at 120,000 rpm) were taken under simulated normal wet-cutting environments, and the consequent damage to the tooth tissue was observed as it occurred. Concentrically engineered bur types produced a superior quality cut surface at the entry, exit, and advancing front aspects of a cavity, as well as less subsurface cracking. Imaging of the coolant water film local to recent cutting operations showed regular spherical cutting debris of 6 to 18 um diameter from the concentric tools, whereas the less-well-engineered burs produced ragged, irregular chips, with 25-40 um diameter debris, indicating far more aggressive cutting actions. This study has shown that there is reduced substrate damage with high-concentricity carbide burs.

Diopside has been developed for use in dental root implants and for the filling of bone defects. In previous studies, diopside developed hydroxyapatite (HA) on its surface and achieved a direct bond with bone. The purpose of this study was to investigate the mechanism of crystal formation on the diopside surface. We ultrastructurally evaluated the interface between new diopside-induced crystals and diopside. Specimens were prepared in three experiments: (1) Granular diopside was immersed in simulated body fluid (SBF); (2) granular diopside was implanted into a cavity in rabbit bone; and (3) a diopside dental root implant was implanted into a Japanese monkey. The specimens were examined by contact microradiography, high-resolution transmission electron microscopy, and analytical electron microscopy. In the experiment with SBF, many platelet-like crystals formed in the diopside surface layer. The lattice of diopside and that of the new crystals were very close, but no clear continuation of the lattice was observed. In the experiments which used a rabbit and a monkey, contact microradiography showed close contact between bone and diopside. High-resolution transmission electron microscopy revealed crystal growth from the diopside surface layer, and continuity between the diopside lattice and that of the new crystals. The morphological characteristics of the new crystals and the results of these analyses suggest that these new crystals are HA. With regard to the mechanism by which crystals are formed on the diopside surface layer, it is possible that epitaxial crystal growth could originate as a nucleus on the surface. In this case, epitaxial crystal growth of primarily octacalcium phosphate (OCP) may have occurred, and this may have changed to HA by a phase transition. However, epitaxial growth of OCP on the diopside surface is still highly speculative, since there is no direct supporting evidence.

Acid phosphate is one of the major impurities incorporated into bioapatites, and its quantity and environment in forming mineral have been used as diagnostic probes to pursue acidic precursor(s). Currently, little is known about the structural feature of non-stoichiometric octacalcium phosphate (OCP), which has been advocated to be, most plausibly, mineral salt initially formed during amelogenesis. In the present report, we attempt to define the state of acid phosphate in OCP crystals which were Ca-deficient and contained 40% total phosphate as acid phosphate. We assessed fractions of acid phosphate in discrete environments by extracting the crystals in either deionized water, 10 mmol/L NaOH solution (initial pH 11), or 150 mmol/L Tris buffer at pH 7.4. Solid samples before and after the treatments were examined by chemical analyses and x-ray diffraction. The results indicated that successive extractions with use of the alkaline solution brought about a reversible change (not hydrolysis) in the interior structure of OCP, which accompanied a marked decrease in acid phosphate. A substantial part of the lost acid phosphate was restored during subsequent treatments at neutral pH, and, intriguingly, this restoration accompanied a re-ordering of OCP structure. The data suggested that the acid phosphate in OCP is separated into three pools: (a) a stable pool corresponding to roughly 50 to 60% of the total acid phosphate, (b) a reversibly exchangeable pool corresponding to 25 to 30% of the acid phosphate which may exist either in the water layer or on crystal surfaces, and (c) an unstable (or irreversibly lost) pool corresponding to 15 to 20% of the acid phosphate, a part of which might be explained by the presence of excess hydrogen in OCP. The present work supports the concept that protons and, to a lesser magnitude, phosphate species can diffuse into and out of the OCP lattice prior to initiation of its hydrolytic transition into apatite.

We recently demonstrated the advantages of back-scattered electron images (COMPO) in the visualization of dentinal caries, and the relationship of the change in the dentin fluorescence pattern in caries lesions. However, the exact nature of these changes is not known. In this paper, the nature of the changes in the areas with reduced mineral content in COMPO images was investigated. We examined the relation of changes in mineral elements and the appearance of soft carious and sound dentin in COMPO images using a scanning electron microscope (SEM) equipped with an electron probe microanalyzer (EPMA). Rat molars with small dentinal caries lesions just under the DEJ were chosen for the study. The Ca, P, Na, Mg, Zn, F, and total contents were determined by EPMA from five different dentin sites, and the Ca/P and Mg/Ca ratios were calculated. Generally, the lowest contents were found in caries lesions and highest in mantle dentin, with the exceptions of Mg and Zn. The Ca/P ratio was lowest in mantle dentin and highest in carious dentin. The results confirm that the change in fluorescence in the dentinal caries lesion is correlated with the very initial changes in mineral content, and that EPMA used in combination with COMPO images is a useful tool for determining small changes in mineral elements in the carious and adjacent areas of dentin.

The proton probe has been used to map F concentration changes in the enamel of 15 teeth showing clinical evidence of caries. Thin sections through the lesions were microradiographed and measurements made of the surface zone (radiodense) and body (radiolucent) areas. Each section was then scanned with a focused beam of 2.5 MeV protons, 2000 spot analyses being performed over areas up to 2 x 3 mm. F was determined by detecting y rays from a nuclear reaction and the data used to construct 3-D surface plots. The maximum F concentration in the lesion surface zone was extremely variable, ranging from 1750 to 21,700 ppm, and rarely occurred over the deepest part of the lesion. F levels were elevated in the lesion body but usually to a small extent only. A large increase in F throughout the lesion body was found in 3 lesions only, and was associated with a surface zone that was thin or of low x-ray density. Relatively small F increases in the lesion body were associated with either a thick, x-ray dense surface layer having a greatly increased F level (> 10,000 ppm) or, conversely, with a surface layer having a relatively small F increase. Since F uptake can be regarded as a "marker" of past remineralization events, this study shows that remineralization can and does occur in the body of natural enamel caries lesions, especially when the surface layer is thin or lost. Fluoride availability that encourages the formation of an extremely dense surface layer may result in under-achievement of this natural repair process in the lesion body.

To investigate the mechanism of collagen accumulation in oral submucous fibrosis (OSF) tissues, we examined the biosynthesis of collagen in fibroblast cultures established from OSF lesions. Fibroblasts obtained from four of ten OSF specimens showed more than a 1.5-fold increase in the production of collagens compared with fibroblasts from age-, sex-, and passage-matched normal controls (p < 0.05). When the relative amounts of collagen synthesis were estimated by SDS polyacrylamide gel electrophoresis, it was found that both OSF and control cells produced about 85% type I collagen and 15% type III collagen. The ratio of α1(I) to a2(I) chains was about 3:1 in OSF cells instead of the 2:1 expected for type I collagen. The excess al(I) chains could mean that collagen type I trimer was synthesized by the fibroblasts. These findings suggest that collagen overproduction and a reduced degradation of the structure-stable collagen type I trimer synthesized by OSF fibroblasts might contribute to the accumulation of collagen in OSF lesions in vivo. The mechanism(s) of increased procollagen production were analyzed by Northern blot, slot blot, and Southern blot. The OSF fibroblast strains with elevated collagen production also contained higher-than-normal levels of procollagen mRNA, and the ratios of α1(I), a2(I), and α1(III) procollagen mRNAs were compatible with the results of corresponding procollagen a chains. The gene copy number of proa2(I) collagen gene in OSF fibroblasts was about 1.05. No gene amplification was found. These results indicate that expression of these procollagen genes in cultured fibroblasts is regulated at the transcriptional level.

The primary ecological niche for suspected periodontal pathogens seems to be the subgingival area, even though periodontal pathogens are also frequently recovered from saliva. The interrelationship of different periodontal conditions and the salivary levels of suspected periodontal pathogens is not known. In the present study, salivary levels of Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, Campylobacter rectus, and Peptostreptococcus micros were determined by bacterial culture and related to clinical periodontal status in 40 subjects with either advanced, moderate, or initial/no periodontitis. Culture-positive subjects harbored the 5 bacterial species in mean numbers ranging from 2 x 105 to 6 x 107 colony-forming units (CFU)/mL saliva. A. actinomycetemcomitans was found in none and P. gingivalis in one of the subjects with initial periodontitis, whereas both species were found in 33% and 44%, respectively, of the subjects with moderate periodontitis and in 60% and 40%, respectively, of the subjects with advanced periodontitis. The mean numbers of CFU/mL of P. intermedia, C. rectus and P. micros were significantly higher in subjects with advanced periodontitis than in subjects with initial/no periodontitis. Ten patients with advanced periodontitis were treated mechanically and with adjunctive systemic metronidazole, and were re-examined 1 and 6 months after treatment. Periodontal treatment eradicated or significantly reduced the levels of salivary periodontal pathogens for half a year, whereas in untreated subjects, the levels and the detection frequencies generally remained fairly stable. In conclusion, the results showed that the salivary levels of periodontal pathogens reflect the periodontal status of the patient.

The identification of periodontal pathogens by conventional methods is time-consuming and difficult. Therefore, a multiplex PCR method for simultaneous detection of Actinobacillus actinomycetemcomitans (A.a.) and Porphyromonas gingivalis (P.g.) was developed for rapid and easy determination of these risk-indicator bacteria in human periodontal disease. The PCR primers were designed to hybridize to various regions of 16S rRNA genes, and a hot-start technique was used to obtain maximum sensitivity and specificity. This method can detect both of these bacteria in subgingival plaque samples at concentrations as low as 5 to 50 cells per sample. The sensitivity, however, was even 10 times better when the bacteria were analyzed in a water suspension. Since the only step between sample collection and the actual analysis is a brief centrifugation of the patient sample, the detection can be readily carried out in four hours. The performance of the method was studied with 36 patient samples. The results showed that the PCR method detected A.a. (44% vs. 25%, respectively) and P.g. (56% vs. 42%, respectively) more often than the conventional culture in plaque samples. Thus, our multiplex PCR method is rapid and more effective than conventional protocols in detecting these periodontal pathogens.

Gingival fibroblasts function as accessory immune cells and are capable of synthesizing cytokines in response to lipopolysaccharides (LPS) from Gram-negative microbes. Recently, we have isolated, cloned, and characterized two cell lines which exhibit characteristics of periodontal ligament (PDL) cells. In this report, we demonstrate that PDL cells showing osteoblast-like phenotype are not LPSresponsive cells. However, treatment of PDL cells with tumor necrosis factor-α (TNF-a) inhibits the expression of their osteoblast-like characteristics. As a consequence of this TNF-a-induced phenotypic change, PDL cells become LPSresponsive, i.e., synthesize several pro-inflammatory cytokines in response to LPS. These phenotypic changes occur at concentrations of TNF-a that are frequently observed in tissue exudates during periodontal inflammation, suggesting a physiological significance for these in vitro observations. It is of interest that TNF-a-induced phenotypic changes in PDL cells are transient, since removal of rhTNF-a from the supernatants of PDL cell cultures results in re-acquisition of the osteoblast-like characteristics and lack of LPS responsiveness of PDL cells. These results suggest that TNF-a, by regulating the PDL cell functions, may allow these cells to participate in the disease process as accessory immune cells at the expense of their structural properties.

The involvement of transglutaminase C (TGase C) in morphogenesis and cytodifferentiation during glandular tubule formation was addressed by immunolocalization of the protein at different stages of prenatal human submandibular gland development in 100 fetuses and 20 adult salivary glands. Immunocytochemical detection was carried out using a monospecific antibody to TGase C. The results showed TGase C reactivity in both acini and ducts early in development (from 10 to 14 weeks), followed by a marked increase in ductal activity and a decline in acinar activity up to 32 weeks. During the peak of reactivity at 25 to 32 weeks, staining was concentrated in the apical ends of the columnar cells. In the adult, staining was weakly and diffusely distributed in the striated and excretory ducts. Western blot analysis of the cellular extracts of pooled samples from various stages of salivary gland development showed a single strong band at 76 kDa early in development. This band became weaker after 32 weeks of prenatal development and in the adult. These findings of transient high expression of TGase C, which coincide with the development of tubulo-alveolar structure, suggest that TGase C may play a role in morphogenesis in human salivary gland development.