- Email: [email protected]
A double blind randomised controlled clinical trial comparing a novel anti-stain and calculus reducing dentifrice with a standard fluoride dentifrice
journal of dentistry 41 (2013) 313–320
Available online at www.sciencedirect.com
journal homepage: www.intl.elsevierhealth.com/journals/jden
A double blind randomised controlled clinical trial comparing a novel anti-stain and calculus reducing dentifrice with a standard fluoride dentifrice Adrian K. Jowett a,*, Ian Marlow b, Andrew Rawlinson a a b
Academic Unit of Restorative Dentistry, School of Clinical Dentistry, University of Sheffield, Sheffield, S10 2TA, UK Boots Brands and Exclusives, Boots UK, 1 Thane Road, Nottingham, NG90 1BS, UK
article info
abstract
Article history:
Objectives: This clinical trial tested the anti-stain efficacy at 3 and 6 months of a novel,
Received 9 November 2011
sodium polyaspartate-containing, anti-stain dentifrice. In addition, the efficacy of the new
Received in revised form
dentifrice in controlling gingival inflammation and inhibition of calculus deposition was
1 November 2012
tested.
Accepted 10 December 2012
Methods: Participants were recruited to this double blind randomised control clinical trial, and allocated to either test or control groups. The presence of stain and calculus were entry criteria. Measurements of stain, calculus and gingival inflammation were recorded using the
Keywords:
Shaw and Murray Stain score, Volpe-Manhold Calculus score and the Modified Gingival
Toothpaste
Index respectively. Measurements were made at baseline, prior to the removal of stain and
Sodium polyaspartate
calculus, and after 3 and 6 months. Missing data were imputed by and the outcomes were
Dental calculus
analysed using univariate analysis.
Gingivitis
Results: At three months, toothpaste containing sodium polyaspartate was better (difference of mean 1.13 with SEM 0.57) than control for the control of dental stain ( p < 0.05). Stain scores also showed a trend in favour of the test product (difference of mean 1.03 with SEM 0.78) at six months ( p > 0.05). There was no difference between toothpastes with respect to calculus deposition or gingival inflammation. Conclusions: Toothpaste containing sodium polyaspartate was more effective than a control toothpaste at preventing deposition of dental stain for 3 months after professional tooth cleaning but showed no significant effect at 6 months. Clinical relevance: Sodium polyaspartate toothpaste was more effective than a control toothpaste at preventing dental stain formation and maybe helpful in controlling staining between episodes of scaling and polishing. # 2012 Elsevier Ltd. All rights reserved.
1.
Introduction
Dental stain and the formation of calculus are of clinical and economic importance. Dental stain can be unsightly, and calculus, typically forming around the necks of teeth, becomes
a rough, hard surface which retains the dental plaque biofilm. This leads to further calculus formation and plaque directly affects the gingival tissues inducing gingivitis that may lead to more serious periodontitis,1 which may ultimately lead to tooth loss. Once dental calculus becomes hard, it may require professional removal. The cost of scaling and polishing teeth
* Corresponding author. Tel.: +44 0114 271 7923; fax: +44 0114 2265484. E-mail address: [email protected] (A.K. Jowett). 0300-5712/$ – see front matter # 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jdent.2012.12.005
314
journal of dentistry 41 (2013) 313–320
to remove dental stain and calculus was at least £140 million to the UK NHS general dental services alone,2 excluding treatment in hospitals and by private contract. Dentifrices reducing stain and calculus formation may therefore lead to improvements in oral health and reduce the demand and cost of dental care. Tooth staining is a common dental complaint and may be controlled to varying degrees by the home use of a dentifrice or by a dental professional undertaking scaling and polishing. In a review of the literature, Watts and Addy3 found the extrinsic staining of teeth to be caused by either the basic colour of compounds incorporated into the dental pellicle (direct staining) or as a result of chemical interaction at the tooth surface (indirect staining). Examples of the former include direct staining from tobacco, tea and coffee, though these and other substances including mouthrinses, medicaments and chromogenic bacteria may also indirectly cause tooth staining. Indirect extrinsic staining can be associated with the use of chlorhexidine mouthrinse or metallic salts within dietary supplements, medications or from occupational exposure to metal salts. Since the removal of stain by dental professionals is costly and labour intensive, there has been considerable interest in its control by the use of dentifrices formulated for this purpose. Whilst the results of some clinical studies are encouraging,4–8 there continues to be a demand for new formulations of dentifrice that aim to improve the control of dental stain, control calculus formation and maintain tooth whiteness. There has also been considerable clinical research on anticalculus agents and the efficacy of these in dentifrices. Fairbrother and Heasman13 reviewed these studies and concluded there was evidence for some formulations effectively reducing calculus formation. However, there was a lack of evidence from comparative studies to show that any one formulation was clearly better than the others tested. In their paper, the strategies for controlling calculus formation were also reviewed and included measures to alter the attachment of calculus to tooth surfaces, the prevention of plaque formation and also the inhibition of mineralisation. More recent studies have continued to report the efficacy of new formulations of toothpaste.9–12 A wide variety of agents have been used to reduce calculus formation, including the use of heavy metals to inhibit mineralisation, bisphosphonates, pyrophosphates, polymers and co-polymers that are familiar ingredients in anticalculus dentifrices. Their efficacy in reducing calculus formation in clinical trials varied widely from as little as 9% to as much as 69% in the data they presented, when anticalculus dentifrices were compared with a sodium monofluorophosphate control dentifrice. However, in most studies, anticalculus agents reduced the amount of calculus by approximately 30–50%. The variability may in part be accounted for by different study designs. Sowinski et al.,14 identified some of these variations, and recommended reporting the importance of ensuring participants’ ability to form calculus during a pre-trial phase, scaling prior to the test phase, random assignment to test groups and for studies to be undertaken for periods of up to 6 months. Further research on new agents is warranted to identify better anticalculus agents. One such new agent is sodium polyaspartate. Polyaspartate is an anionic polymer that is used
industrially to inhibit crystal growth and which could be used to reduce dental calculus formation.15,16 In addition, polyaspartate was effective in controlling the development of dental plaque and the formation of a bacterial biofilm when used in a dentifrice formulation in vitro.15 The benefits of polyaspartate both in reducing the build-up, and in increasing the removal of, stain on extracted teeth and on artificial substrates, as well as in reducing calculus crystal growth have been shown in laboratory studies. The aim of this study was to assess the efficacy of a sodium polyaspartate-containing toothpaste compared to a sodium monofluorophosphate control toothpaste in its ability to prevent deposition of dental stain and calculus in the medium term. The working hypothesis was that a sodium polyaspartate-containing toothpaste is more effective than a control in preventing deposition of dental stain for 3 and 6 months after professional tooth cleaning. Secondary outcomes of reduction in deposition of dental calculus and gingivitis were also assessed.
2.
Methods
The research project was independently approved by the South Sheffield Research Ethics Committee (reference STH14105/SSREC 96/308) and conducted in accordance with the World Medical Association Declaration of Helsinki.
2.1.
Sample
Using the data for a negative control dentifrice,14 it was calculated that a sample size of 40 per group would be required to detect a 35% difference between test and control dentifrices, as might be anticipated since in most studies, anti-calculus agents reduce the amount of calculus by between 30 and 50%,13 with an 80% power with respect to calculus formation. It was calculated that fewer participants would be required to detect differences in stain formation based on previous studies conducted within our group. Therefore a sample size of 40 per group was selected. Participants were recruited from staff of the University of Sheffield and Sheffield Teaching Hospital NHS Foundation Trust, by use of email networks. Participants were screened according to the inclusion and exclusion criteria (Table 1). Screening and recruitment was undertaken by AKJ or AR.
2.2.
Toothpaste allocation
Prior to recruitment an allocation sequence according to gender and smoking status, in accordance with a random permitted blocks with strata17 was generated by Dr Robinson of the Statistical Services Unit, The University of Sheffield. Ninety-two participants were recruited to allow for over 10% drop-out, given a consecutive number and allocated by a dental nurse to the test or control group according to the predefined sequence. The test and control toothpastes (composition summarised in Table 2) were packaged identically except for a coded identification number. The coding for the toothpaste was withheld from the clinicians until data analysis was complete.
journal of dentistry 41 (2013) 313–320
315
Table 1 – Inclusion and exclusion criteria for the study. Inclusion criteria
Exclusion criteria
Aged 18 years and over Systemically healthy At least 20 teeth including 33 to 43 BPE 2 i.e. calculus detected on at least one tooth Visible stain on at least 4 incisors
Pregnancy Antibiotics taken within 1 month prior to the study Presence of oral prostheses or orthodontic appliances BPE > 3 Currently receiving dental treatment that would result in the removal of plaque or calculus
Available for 6 months after recruitment to the study
2.3.
Outcome assessment
The primary outcome measure was extrinsic staining of the upper and lower incisors with secondary outcomes of the presence of calculus and gingival inflammation. Clinical examinations were undertaken by AKJ and AR. Every attempt was made for all the measurements for a given participant to be undertaken by the same clinical examiner on each occasion during the trial. The only exception permitted, was the attendance of a participant when their allocated examiner was unavailable. At each attendance, extrinsic stain on the labial and palatal/lingual surfaces of the upper and lower central and lateral incisors was scored using the Shaw & Murray Stain Index.18 Calculus formation was measured on the lingual aspects of 6 mandibular anterior teeth using the VolpeManhold Calculus Index19 score. The Modified Gingival Index (MGI) was used to assess gingival health.20 The gingival margins and interdental papillae of the Ramfjord index teeth were examined and scored.
2.4.
Interventions
Assessment, recruitment and clinical interventions were undertaken in the Periodontology Clinic in the Department
of Restorative Dentistry, Charles Clifford Dental Hospital, Sheffield, between 1st October 2005 and 31st August 2006. After recruitment, the baseline data were recorded and each participant was given a thorough scale using ultrasonic and hand instruments followed by polishing with prophylaxis paste (Sultan Topex Medium, Henry Schein, Bradford, England). Standard advice on tooth brushing (Modified Bass technique twice daily) and the use of the toothpaste (volume of a small pea, rinse and do not swallow) was provided. The use of inter-dental cleaning aids was discouraged and the use of mouthwashes or other toothpastes prohibited for the duration of the study. Three tubes of test or control toothpaste, according to the allocation scheme, were given to each participant who was masked to their allocation. Participants were provided with additional toothpaste on demand. Adherence to the trial was monitored by asking participants to mark a calendar each time they brushed their teeth with the allocated toothpaste. Participants were re-examined at 3 and 6 months. At the conclusion of the trial, participants were offered a scale and polish. At each attendance, participants were asked to report any problems or adverse events and these were recorded anonymously with the experimental data.
2.5. Table 2 – Composition of the control and test dentifrices as % by weight. Control Aqua Hydrated silica Sorbitol PEG-32 Cellulose gum Sodium bicarbonate Aroma Sodium lauryl sulfate Sodium benzoate Sodium saccharin Sodium fluoride Limonene CI 77891
Test 58.38 17.50 17.50 1.39 1.20 1.00 0.79 0.75 0.50 0.26 0.22 0.01 0.50
Aqua Hydrated silica Cellulose
26.50 6.39 10.00
Cellulose gum Sodium bicarbonate Aroma
1.00 5.00 0.97
Sodium saccharin Sodium fluoride Limonene CI 77891 Glycerin Pentasodium triphosphate Sodium polyaspartate 2000–3000 g/mol Disodium phosphate Cocamidopropyl betaine Sodium phosphate Zinc citrate Bromochlorophene
0.26 0.32 0.03 0.75 39.92 3.00 2.00 1.00 1.30 1.00 0.46 0.10
Reliability of measurements
Fifteen participants had repeat measurements made of the parameters recorded by both clinical examiners (AR and AKJ), to determine the inter and intra-examiner reliability. Repeat measurements were performed by examiners at least one hour after the first measurements and were undertaken blind to them. Intra-examiner repeatability assessment of AR gave R = 0.65, 0.95 and 0.85 for mean stain, mean calculus index and mean MGI respectively. For AKJ the equivalent values were R = 0.79, 0.94 and 0.90. These data indicate that both examiners show substantial to excellent repeatability of measurements. For inter-examiner reproducibility R = 0.60, 0.92 and 0.71 for mean stain, mean calculus index and mean MGI respectively, indicating moderate to excellent reproducibility of measurements.
2.6.
Data analysis
Data were entered onto Excel spreadsheets using the participant code to maintain the anonymity of the participants, and then transferred to the PASW18.0.3 (formerly SPSS, IBM Corporation, Somers, NY 10589, USA) statistical package for
316
journal of dentistry 41 (2013) 313–320
detailed analysis. The participants’ age, gender and domicile postcodes were obtained from hospital records. The GeoConvert tool (http://geoconvert.mimas.ac.uk/) was used to link participants’ postcodes with their Index of Multiple Deprivation 2007 (IMD) score21 as a measure of socioeconomic status. IMD uses 38 indicators to assess access to resources for 32 482 areas in England. The culmulative impact of income, health, education and other domains can be used for Government planning but has been shown to relate to health in both rural and urban areas.22 Histograms confirmed the data were normally distributed before further analysis. All statistical tests of hypotheses were two-sided, and employed a level of significance of a = 0.05 at 95% confidence level. The intra- and inter-examiner data were analysed by calculation of the intraclass correlation coefficient (ICCC). Stain, calculus score and MGI scores were compared at 3 and 6 months, using a generalised linear model analysis with Wald’s Chi-square test for hypothesis testing using the follow-up data as the dependent variable, the dentifrice as the fixed factor and the baseline scores as the covariate. These analyses were undertaken on an intention to treat basis using both principled (multiple imputation) and unprincipled (last observation carried forward) methods for imputating missing data.
2.7.
Multiple imputation
Little’s test was used to demonstrate that the missing data were missing completely at random (MCAR p = 0.8) and thus amenable to imputation.23 For principled imputation a regression model was utilised using the complete baseline covariate dataset as predictor and longitudinal data as both predictor and to be imputed where missing. A constraint was applied that all imputed data must be 0 as the clinical indices cannot be negative. The three outcome measures were imputed independently. 20 data sets were imputed for each outcome as preliminary analyses showed that with this number of iterations the means tended to a constant value (data not shown). Means and standard error of the means were obtained of the pooled original and imputed data sets.
Fig. 1 – Flow of participants through the trial.
3.1.
As shown in Table 3, the two cohorts were similar with respect to demographic characteristics.
3.2. 2.8.
Adverse events
Last observation carried forward
A widely adopted unprincipled imputation method is last observation carried forward (LOCF) and this was done by substituting zero for missing stain and calculus indices and baseline values for gingival index.
3.
Demographic data
At 3 month review one participant in each of the test (1/38) and control (1/43) groups reported that ulcers were lasting longer than usual. No pathology was visible on examination. Some participants (18/92) commented upon the ‘dull’ or ‘bland’ taste of both of the dentifrices.
Results
Ninety-two participants were recruited to the study with 46 in each of control and test groups. At 3 months, 11 participants (3 control and 8 test) had withdrawn from the study leaving 81 participants available for data analysis. At the end of the study a further 3 control participants had withdrawn leaving 78 participants available for data collection (40 control and 38 test) as detailed in Fig. 1.
Table 3 – Summary demographic descriptive data showing gender, age and Index of Multiple Deprivation 2007 score21 for the control and test groups.
Gender Age IMD 2007
Control
Test
19 male, 27 female Range 28–64 years, x¯ = 43.3 x¯ = 18.6, SD = 14.9
20 male, 26 female Range 21–59 years, x¯ = 37.5 x¯ = 18.3, SD = 13.3
317
journal of dentistry 41 (2013) 313–320
Table 4 – Mean values of clinical measurements at baseline and imputed datasets at 3 and 6 months with 1 standard error of the mean in parentheses. Extrinsic stain
Baseline 3 month 6 month
3.3.
Calculus index
Test
Control
Test
Control
Test
Control
4.88 (0.53) 3.04 (0.41) 5.11 (0.55)
5.43 (0.52) 4.17 (0.40) 6.14 (0.55)
0.64 (0.044) 0.33 (0.034) 0.40 (0.038)
0.71 (0.058) 0.38 (0.031) 0.44 (0.038)
0.62 (0.059) 0.33 (0.023) 0.32 (0.022)
0.65 (0.052) 0.33 (0.023) 0.27 (0.021)
Baseline comparisons
At baseline, the data for all parameters were normally distributed and the two groups were similar (Table 4 and Fig. 2a–c).
3.4.
Mean gingival index
Hypothesis testing from imputed data sets
Means and standard errors of the means for both baseline and imputed datasets are presented in Table 4 and Fig. 2a–c. All
three clinical measurements were markedly reduced 3 months after professional cleaning at baseline. Extrinsic staining and calculus showed an increase between 3 and 6 months whereas the mean gingival index maintained its reduced level. Table 5 summarises p values from hypothesis testing on the imputed datasets. MI and LOCF show a statistically significantly greater effect of the test dentifrice on preventing dental staining at 3 months.
4.
Discussion
4.1.
Relative efficacy of dentifrices
The aim of this double-blind randomised controlled trial was to test the efficacy of a polyaspartate-containing dentifrice in controlling deposition of dental stain for 3 and 6 months. The novel dentifrice effectively reduced development of stain compared to a control fluoridated toothpaste at 3 months and the difference between the dentifrices was the same magnitude but not significant at 6 months. The dentifrices were similarly effective with respect to their ability to control calculus deposition and gingival inflammation. The Relative Dentine Abrasivity scores for the test and control were 48.1 and 38.5 respectively, which while not matched, indicate that the dentifrices both have relatively low abrasivity by comparison with proprietary ‘whitening’ toothpastes where scores over 100 are commonplace.24 The apparent diminution of effect at 6 months appears to be caused by three features of the study design, whose effects on the analysis are culmulative: (a) At 3 months the test group had adjusted mean stain scores around 20% lower than the controls group and 5% lower adjusted mean calculus scores. These effects are substantially smaller than those predicted when the sample size was estimated before the trial and show that a larger
Fig. 2 – (a) Extrinsic stain scores (shown as MI mean W 1 S.E.M.) in test and control dentifrices at baseline, 3 and 6 month review (* indicates a statistically significant difference between the dentifrices with p < 0.05 with baseline scores as a covariate). (b) Calculus scores (shown as MI mean W 1 S.E.M.) in test and control dentifrices at baseline, 3 and 6 month review. (c) Mean Gingival Index (shown as MI mean W 1 S.E.M.) in test and control dentifrices at baseline, 3 and 6 month review.
Table 5 – Summary of hypothesis testing on imputed datasets using Wald’s Chi-squared test. Extrinsic stain
Calculus index
Mean gingival index
Multiple imputation 3 month p = 0.04 p = 0.19 6 month
p = 0.27 p = 0.50
p = 0.90 p = 0.08
Last observation carried forward p = 0.01 3 month 6 month p = 0.09
p = 0.27 p = 0.54
p = 0.90 p = 0.35
318
journal of dentistry 41 (2013) 313–320
sample is required to reveal significant differences. Previous studies of the relative effectiveness of test dentifrices in either preventing staining de novo or removing pre-existing staining show substantial differences. For removal of forced stain, Moran et al.,25 found negligible benefit from a test dentifrice; whereas Isaacs et al.,26 found that 3 dentifrices inhibited all stain development for 3 months. (b) The trial suffered from substantial loss to follow-up with a resultant loss of power during ITT analysis. It is recommended for future trials that loss to follow-up is minimised by making rigorous attempts to maintain contact with participants. (c) There is a greater variance at 6 months due to a combination of a greater proportion of imputed data and an inevitable increase in variance of clinical data. Dental staining is thought to arise through a combination of three processes. First, by complexation of metal ions, such as those of copper, iron and tin into the pellicle. This may be brought about by agents within mouthwashes or by plaque microorganisms.27 Polyaspartate has been shown to reduce bacterial adhesion to hydroxyapatite in vitro.15 As yet the mechanism for this is unknown but it is likely that potent binding between aspartate and the mineral phase restricts bacterial binding to the mineral15 by competing with extracellular adhesion proteins. It might be expected that such competition may lead to a quantitative reduction in plaque on tooth surfaces. However this was not seen in a short in vivo experiment28 although a qualitative change in the plaque cannot be excluded. Fluoropolymers have been shown to reduce binding of mixed bacterial cultures to hydroxyapatite29 but showed only a modest stain inhibiting effect suggesting that other modes of staining are clinically significant. Second, clinically-visible dental stain contains only traces of reduced metals and is primarily a layer of highly calcified organic material.30 Polyaspartate is able to sequester calcium ions and inhibit precipitation of calcium phosphate-containing salts at physiological pH31 and thus it is likely that the reduction in staining seen in this trial may be due to the test dentifrice’s ability to reduce mineralisation. Although sodium polyasparate is known to be able to dissolve hydroxyapatite and brushite32 this effect is greatest at a pH significantly more alkaline than that found in the oral cavity and this will limit the potentially destructive action, as will the short duration of exposure to the dentifrice. Third, many coloured vegetable foodstuffs contain polyphenols which can bind to hydroxyapatite via proline rich salivary proteins.33 Polyaspartates are likely to compete with proline-rich proteins at the tooth surface thereby limiting adherence of polyphenols. This model of extrinsic staining was explored by Shellis et al.,34 who showed that polyphosphates could desorb protein and polyphenolic stains from hydroxyapatite in vitro. Their data indicated that a reduction in staining was maximal when the experimental agent could compete electrostatically with protein at the tooth surface and chelate the calcium salts of chromogens. Both polyphosphate and polyaspartate fulfil these criteria.
In addition to the modes of adhesion outlined above, adherent proteins and blocking agents are subject to degradation by host and bacterial enzymes. Thus the period for which an agent is resident on the tooth surface and able to resist adhesion of dental stain may be quite brief.35 One may therefore expect that dental staining develops progressively and that its rate of clinical appearance is delayed by the polyaspartate-containing dentifrice resulting in its reestablishment at the 6 month examination. The Shaw and Murray Stain Index18 is a technically simple procedure relying on the subjective transfer of the appearance of dental stain onto a grid by the examiner. Other studies have developed more objective assays utilising spectrophotometry, image analysis36 or quantitative light-induced fluorescence (QLF).37 These techniques do not remove subjectivity from assessment of staining as demarcation of areas of interest and positioning of the test teeth is highly dependent on the operator. Nonetheless they have achieved excellent intra-operator repeatability in vitro. However, currently the key barrier to their use is the requirement for accurate alignment of the apparatus in relation to the test teeth. In this clinical trial the intra- and inter-operator repeatability and reproducibility for the Shaw and Murray Stain Index were sufficient.
4.2.
Interpretation of imputation
Ideally all participants in a clinical trial would receive all interventions and be followed up at intervals prescribed by the protocol. However it is inevitable that some will be lost, perhaps as a consequence of the intervention.38 Various statistical methods exist for imputation of missing data but as it is impossible to prove that data is MCAR, none of these is completely satisfactory.38 It has therefore become commonplace for more than one method of imputation to be used, ideally including MI,39 and a conservative judgement made based on the statistical findings. Whilst it is argued that MI is dependent on the quality of model, unprincipled methods all tend to reduce the participant variation in the data.38 It is often believed that LOCF is conservative but in this study LOCF will tend to overestimate the efficacy of dentifrices by carrying forward the zero indices resulting from professional cleaning. Nonetheless the imputation methods converge in finding a statistically significant effect of the test dentifrice on dental staining at 3 months.
4.3.
Outcomes
This trial has shown, in agreement with previous studies, that a dentifrice containing agents that competitively bind to hydroxyapatite will reduce development of dental stain and calculus. The experimental polyaspartate dentifrice had a significant effect on stain formation, compared to the control dentifrice in the 3 months after dental prophylaxis. The experimental dentifrice appears effective at retarding the development of stain on recently cleaned tooth surfaces. However, the efficacy in controlling stain at 6 months should be interpreted with caution due to the differences being clinically small at this interval.
journal of dentistry 41 (2013) 313–320
Conflict of interest statement Professor Rawlinson and Dr. Jowett had full access to all the data in this study and take responsibility for the integrity of the data and the accuracy of the data analysis. Neither of these authors has a commercial or financial interest with Boots UK Ltd., Nottingham, UK. Mr. Marlow of Boots UK Ltd. contributed his experience only to the study design and did not participate in the clinical study. Boots UK Ltd. supplied the research materials and partially funded the clinical costs. This research was supported by research grants from the School of Clinical Dentistry, University of Sheffield and University of Sheffield and Boots National Centre for Applied Research in Oral Health.
Acknowledgements We are grateful to Mrs. S. Witham and Mrs. C. Hagglington for their clinical assistance, Mrs. P. Baker and Mrs. S. Wan for coordinating the research clinics, Dr. D. Robinson, Statistical Services Unit, The University of Sheffield for his advice on study design and Professor P.G. Robinsion, Academic Unit of Dental Public Health, The University of Sheffield for his advice on intention to treat analysis.
references
1. Kinane DF. Causation and pathogenesis of periodontal disease. Periodontology 2000 2001;25:8–20. 2. National Audit Office. Reforming NHS Dentistry: Ensuring effective management of risks. Report by the Comptroller and Auditor General; 25 November 2004. 3. Watts A, Addy M. Tooth discolouration and staining: a review of the literature. British Dental Journal 2001;190: 309–16. 4. Walsh TF, Rawlinson A, Wildgoose D, Marlow I, Haywood A, Ward JM. Clinical Evaluation of the stain removing ability of a whitening dentifrice and stain removing system. Journal of Dentistry 2005;33:413–8. 5. Xie P, Lu J, Wan H, Hao Y. Effect of toothpaste containing dlimonene on natural extrinsic smoking stain: a 4-week clinical trial. American Journal of Dentistry 2010;23:196–200. 6. Prasad KV, Sreenivasan PK, Rajesh G, Ramya K, Rao CB, DeVizio W. The efficacy of dentifrices on extrinsic tooth stains among community dwelling adults in India – a randomised controlled trial. Community Dental Health 2011;28:201–5. 7. Joiner A. Whitening toothpastes: a review of the literature. Journal of Dentistry 2010;38(Suppl. 2):e17–24. [Epub 2010 May 24]. 8. Tere´zhalmy G, Chaves E, Bsoul S, Baker R, He T. Clinical evaluation of the stain removal efficacy of a novel stannous fluoride and sodium hexametaphosphate dentifrice. American Journal of Dentistry 2007;20:53–8. 9. Schiff T, Saletta L, Baker RA, He T, Winston JL. Anticalculus efficacy and safety of a stabilized stannous fluoride/sodium hexametaphosphate dentifrice. Compendium of Continuing Education Dentistry 2005;26(9 Suppl. 1):29–34.
319
10. Winston JL, Fiedler SK, Schiff T, Baker R. An anticalculus dentifrice with sodium hexametaphosphate and stannous fluoride: a six-month study of efficacy. Journal of Contemporary Dental Practice 2007;8:1–8. 11. Schiff T. Anticalculus effect of a cetylpyridinium chloride/ zinc gluconate mucoadhesive gel: results of a randomized, double-blind, controlled clinical trial. Journal of Clinical Dentistry 2007;18:79–81. 12. Llena C, Forner L, Vento C. Anticalculus efficacy of a new dentifrice. Quintessence International 2009;40:497–501. 13. Fairbrother KJ, Heasman PA. Anticalculus agents. Journal of Clinical Periodontology 2000;27:285–301. 14. Sowinski JA, Guido WB, Delores MP, Petrone ME, De Vizio W, Volpe AR. A clinical study to assess the anticalculus efficacy of a new dentifrice containing a special grade of silica (Colgate Total Plus Whitening Toothpaste): a clinical trial on adults. Journal of Clinical Dentistry 2002;13:65–8. 15. Guan YH, Lath DL, deGraaf T, Lilley TH, Brook AH. Moderation of oral bacterial adhesion on saliva coated hydroxyapatite by polyaspartate. Journal of Applied Microbiology 2003;94:456–61. 16. Sikes CS. Inhibition of mineral deposition by phosphorylated and related polyanionic peptides. US Patent 5,051,401; 1991. 17. Pocock SJ. Stratified randomization. Clinical trials: a practical approach. Chichester: Wiley; 1983. 80–89. 18. Shaw L, Murray JJ. A new index for measuring stain in clinical trials. Community Dentistry and Oral Epidemiology 1977;5:116–20. 19. Volpe AR, Manhold JH, Hazen SP. In vivo calculus assessment. Part I. A method and its examiner reproducibility. Journal of Periodontology 1965;36:292–8. 20. Lobene RR, Weatherford T, Ross NM, Lamm RA, Menaker L. A modified gingival index for use in clinical trials. Clinical Preventive Dentistry 1968;8:3–6. 21. Noble M, McLennan D, Wilkinson K, Whitworth A, Barnes H, Dibben C. The English Indices of Deprivation 2007, London: communities and local government. Available at: http:// www.communities.gov.uk/documents/communities/pdf/ 733520.pdf; 2008 [accessed 9.8.11]. 22. Jordan H, Roderick P, Martin D. The index of multiple deprivation 2000 and accessibility effects on health. Journal of Epidemiology and Community Health 2000;58:250–7. 23. Kenwood MG, Carpenter J. Multiple imputation: current perspectives. Statistical Methods in Medical Research 2007;16:199–218. 24. Drisko C. Oral hygiene and periodontal considerations in preventing and managing dentine hypersensitivity. International Dental Journal 2007;57:399–410. 25. Moran J, Claydon MCA, Addy M, Newcombe R. Clinical studies to determine the effectiveness of a whitening toothpaste at reducing stain (using a forced stain model). International Journal of Dental Hygiene 2005;3:25–30. 26. Isaacs RL, Bartizek RD, Owens TS, Walters PA, Gerlach RW. Maintenance of tooth color after prophylaxis: comparison of three dentifrices. Journal of Clinical Dentistry 2001;12:51–5. 27. Addy M, Moran J. Mechanisms of stain formation on teeth, in particular associated with metal ions and antiseptics. Advances in Dental Research 1995;9:450–6. 28. Claydon NC, Addy M, Newcombe R, Moran J. The prevention of plaque re-growth by toothpastes and solutions containing block copolymers with and without polypeptide. Journal Of Clinical Periodontology 2005;32:545–8. 29. Churchley D, Rees GD, Barbu E, Nevell TG, Tsibouklis J. Fluoropolymers as low-surface-energy tooth coatings for oral care. International Journal of Pharmaceutics 2008;352: 44–9. 30. Tantbirojn D, Douglas WH, Ko CC, McSwiggen PL. Spatial chemical analysis of dental stain using wavelength
320
31. 32.
33.
34.
35.
journal of dentistry 41 (2013) 313–320
dispersive spectrometry. European Journal of Oral Science 1998;106:971–6. Sikes CS, Wheeler AP. Regulators of biomineralization. Chemtech 1988;18:620–6. Littlejohn F, Eduardo Sa´ez A, Grant SS. Use of sodium polyaspartate for the removal of hydroxyapatite/brushite deposits from stainless steel tubing. Industrial and Engineering Chemical Research 1998;37:2691–700. Proctor GB, Pramanik R, Carpenter GH, Rees GD. Salivary proteins interact with dietary constituents to modulate tooth staining. Journal of Dental Research 2005;84:73–8. Shellis RP, Addy M, Rees GD. In vitro studies on the effect of sodium tripolyphosphate on the interactions of stain and salivary protein with hydroxyapatite. Journal of Dentistry 2005;33:313–24. Jin Y, Yip H-K. Supragingival calculus: formation and control. Critical Reviews in Oral Biology & Medicine 2002;13:426–41.
36. Lath DL, Smith RN, Guan YH, Karmo M, Brook AH. Measurement of stain on extracted teeth using spectrophotometry and digital image analysis. International Journal of Dental Hygiene 2007;5:174–9. 37. Adeyemi AA, Jarad FD, de Josselin de Jong E, Pender N, Higham SM. The evaluation of a novel method comparing quantitative light-induced fluorescence (QLF) with spectrophotometry to assess staining and bleaching of teeth. Clinical Oral Investigation 2010;14: 19–25. 38. Lachin JM. Statistical considerations in the intent-to-treat principle. Controlled Clinic Trials 2000;21:167–89. 39. Blough DK, Ramsey S, Sullivan SD, Yusen R. The impact of using different imputation methods for missing quality of life scores on the estimation of the cost-effectiveness of lung-volume-reduction surgery. Health Economics 2009;18:91–101.
Available online at www.sciencedirect.com
journal homepage: www.intl.elsevierhealth.com/journals/jden
A double blind randomised controlled clinical trial comparing a novel anti-stain and calculus reducing dentifrice with a standard fluoride dentifrice Adrian K. Jowett a,*, Ian Marlow b, Andrew Rawlinson a a b
Academic Unit of Restorative Dentistry, School of Clinical Dentistry, University of Sheffield, Sheffield, S10 2TA, UK Boots Brands and Exclusives, Boots UK, 1 Thane Road, Nottingham, NG90 1BS, UK
article info
abstract
Article history:
Objectives: This clinical trial tested the anti-stain efficacy at 3 and 6 months of a novel,
Received 9 November 2011
sodium polyaspartate-containing, anti-stain dentifrice. In addition, the efficacy of the new
Received in revised form
dentifrice in controlling gingival inflammation and inhibition of calculus deposition was
1 November 2012
tested.
Accepted 10 December 2012
Methods: Participants were recruited to this double blind randomised control clinical trial, and allocated to either test or control groups. The presence of stain and calculus were entry criteria. Measurements of stain, calculus and gingival inflammation were recorded using the
Keywords:
Shaw and Murray Stain score, Volpe-Manhold Calculus score and the Modified Gingival
Toothpaste
Index respectively. Measurements were made at baseline, prior to the removal of stain and
Sodium polyaspartate
calculus, and after 3 and 6 months. Missing data were imputed by and the outcomes were
Dental calculus
analysed using univariate analysis.
Gingivitis
Results: At three months, toothpaste containing sodium polyaspartate was better (difference of mean 1.13 with SEM 0.57) than control for the control of dental stain ( p < 0.05). Stain scores also showed a trend in favour of the test product (difference of mean 1.03 with SEM 0.78) at six months ( p > 0.05). There was no difference between toothpastes with respect to calculus deposition or gingival inflammation. Conclusions: Toothpaste containing sodium polyaspartate was more effective than a control toothpaste at preventing deposition of dental stain for 3 months after professional tooth cleaning but showed no significant effect at 6 months. Clinical relevance: Sodium polyaspartate toothpaste was more effective than a control toothpaste at preventing dental stain formation and maybe helpful in controlling staining between episodes of scaling and polishing. # 2012 Elsevier Ltd. All rights reserved.
1.
Introduction
Dental stain and the formation of calculus are of clinical and economic importance. Dental stain can be unsightly, and calculus, typically forming around the necks of teeth, becomes
a rough, hard surface which retains the dental plaque biofilm. This leads to further calculus formation and plaque directly affects the gingival tissues inducing gingivitis that may lead to more serious periodontitis,1 which may ultimately lead to tooth loss. Once dental calculus becomes hard, it may require professional removal. The cost of scaling and polishing teeth
* Corresponding author. Tel.: +44 0114 271 7923; fax: +44 0114 2265484. E-mail address: [email protected] (A.K. Jowett). 0300-5712/$ – see front matter # 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jdent.2012.12.005
314
journal of dentistry 41 (2013) 313–320
to remove dental stain and calculus was at least £140 million to the UK NHS general dental services alone,2 excluding treatment in hospitals and by private contract. Dentifrices reducing stain and calculus formation may therefore lead to improvements in oral health and reduce the demand and cost of dental care. Tooth staining is a common dental complaint and may be controlled to varying degrees by the home use of a dentifrice or by a dental professional undertaking scaling and polishing. In a review of the literature, Watts and Addy3 found the extrinsic staining of teeth to be caused by either the basic colour of compounds incorporated into the dental pellicle (direct staining) or as a result of chemical interaction at the tooth surface (indirect staining). Examples of the former include direct staining from tobacco, tea and coffee, though these and other substances including mouthrinses, medicaments and chromogenic bacteria may also indirectly cause tooth staining. Indirect extrinsic staining can be associated with the use of chlorhexidine mouthrinse or metallic salts within dietary supplements, medications or from occupational exposure to metal salts. Since the removal of stain by dental professionals is costly and labour intensive, there has been considerable interest in its control by the use of dentifrices formulated for this purpose. Whilst the results of some clinical studies are encouraging,4–8 there continues to be a demand for new formulations of dentifrice that aim to improve the control of dental stain, control calculus formation and maintain tooth whiteness. There has also been considerable clinical research on anticalculus agents and the efficacy of these in dentifrices. Fairbrother and Heasman13 reviewed these studies and concluded there was evidence for some formulations effectively reducing calculus formation. However, there was a lack of evidence from comparative studies to show that any one formulation was clearly better than the others tested. In their paper, the strategies for controlling calculus formation were also reviewed and included measures to alter the attachment of calculus to tooth surfaces, the prevention of plaque formation and also the inhibition of mineralisation. More recent studies have continued to report the efficacy of new formulations of toothpaste.9–12 A wide variety of agents have been used to reduce calculus formation, including the use of heavy metals to inhibit mineralisation, bisphosphonates, pyrophosphates, polymers and co-polymers that are familiar ingredients in anticalculus dentifrices. Their efficacy in reducing calculus formation in clinical trials varied widely from as little as 9% to as much as 69% in the data they presented, when anticalculus dentifrices were compared with a sodium monofluorophosphate control dentifrice. However, in most studies, anticalculus agents reduced the amount of calculus by approximately 30–50%. The variability may in part be accounted for by different study designs. Sowinski et al.,14 identified some of these variations, and recommended reporting the importance of ensuring participants’ ability to form calculus during a pre-trial phase, scaling prior to the test phase, random assignment to test groups and for studies to be undertaken for periods of up to 6 months. Further research on new agents is warranted to identify better anticalculus agents. One such new agent is sodium polyaspartate. Polyaspartate is an anionic polymer that is used
industrially to inhibit crystal growth and which could be used to reduce dental calculus formation.15,16 In addition, polyaspartate was effective in controlling the development of dental plaque and the formation of a bacterial biofilm when used in a dentifrice formulation in vitro.15 The benefits of polyaspartate both in reducing the build-up, and in increasing the removal of, stain on extracted teeth and on artificial substrates, as well as in reducing calculus crystal growth have been shown in laboratory studies. The aim of this study was to assess the efficacy of a sodium polyaspartate-containing toothpaste compared to a sodium monofluorophosphate control toothpaste in its ability to prevent deposition of dental stain and calculus in the medium term. The working hypothesis was that a sodium polyaspartate-containing toothpaste is more effective than a control in preventing deposition of dental stain for 3 and 6 months after professional tooth cleaning. Secondary outcomes of reduction in deposition of dental calculus and gingivitis were also assessed.
2.
Methods
The research project was independently approved by the South Sheffield Research Ethics Committee (reference STH14105/SSREC 96/308) and conducted in accordance with the World Medical Association Declaration of Helsinki.
2.1.
Sample
Using the data for a negative control dentifrice,14 it was calculated that a sample size of 40 per group would be required to detect a 35% difference between test and control dentifrices, as might be anticipated since in most studies, anti-calculus agents reduce the amount of calculus by between 30 and 50%,13 with an 80% power with respect to calculus formation. It was calculated that fewer participants would be required to detect differences in stain formation based on previous studies conducted within our group. Therefore a sample size of 40 per group was selected. Participants were recruited from staff of the University of Sheffield and Sheffield Teaching Hospital NHS Foundation Trust, by use of email networks. Participants were screened according to the inclusion and exclusion criteria (Table 1). Screening and recruitment was undertaken by AKJ or AR.
2.2.
Toothpaste allocation
Prior to recruitment an allocation sequence according to gender and smoking status, in accordance with a random permitted blocks with strata17 was generated by Dr Robinson of the Statistical Services Unit, The University of Sheffield. Ninety-two participants were recruited to allow for over 10% drop-out, given a consecutive number and allocated by a dental nurse to the test or control group according to the predefined sequence. The test and control toothpastes (composition summarised in Table 2) were packaged identically except for a coded identification number. The coding for the toothpaste was withheld from the clinicians until data analysis was complete.
journal of dentistry 41 (2013) 313–320
315
Table 1 – Inclusion and exclusion criteria for the study. Inclusion criteria
Exclusion criteria
Aged 18 years and over Systemically healthy At least 20 teeth including 33 to 43 BPE 2 i.e. calculus detected on at least one tooth Visible stain on at least 4 incisors
Pregnancy Antibiotics taken within 1 month prior to the study Presence of oral prostheses or orthodontic appliances BPE > 3 Currently receiving dental treatment that would result in the removal of plaque or calculus
Available for 6 months after recruitment to the study
2.3.
Outcome assessment
The primary outcome measure was extrinsic staining of the upper and lower incisors with secondary outcomes of the presence of calculus and gingival inflammation. Clinical examinations were undertaken by AKJ and AR. Every attempt was made for all the measurements for a given participant to be undertaken by the same clinical examiner on each occasion during the trial. The only exception permitted, was the attendance of a participant when their allocated examiner was unavailable. At each attendance, extrinsic stain on the labial and palatal/lingual surfaces of the upper and lower central and lateral incisors was scored using the Shaw & Murray Stain Index.18 Calculus formation was measured on the lingual aspects of 6 mandibular anterior teeth using the VolpeManhold Calculus Index19 score. The Modified Gingival Index (MGI) was used to assess gingival health.20 The gingival margins and interdental papillae of the Ramfjord index teeth were examined and scored.
2.4.
Interventions
Assessment, recruitment and clinical interventions were undertaken in the Periodontology Clinic in the Department
of Restorative Dentistry, Charles Clifford Dental Hospital, Sheffield, between 1st October 2005 and 31st August 2006. After recruitment, the baseline data were recorded and each participant was given a thorough scale using ultrasonic and hand instruments followed by polishing with prophylaxis paste (Sultan Topex Medium, Henry Schein, Bradford, England). Standard advice on tooth brushing (Modified Bass technique twice daily) and the use of the toothpaste (volume of a small pea, rinse and do not swallow) was provided. The use of inter-dental cleaning aids was discouraged and the use of mouthwashes or other toothpastes prohibited for the duration of the study. Three tubes of test or control toothpaste, according to the allocation scheme, were given to each participant who was masked to their allocation. Participants were provided with additional toothpaste on demand. Adherence to the trial was monitored by asking participants to mark a calendar each time they brushed their teeth with the allocated toothpaste. Participants were re-examined at 3 and 6 months. At the conclusion of the trial, participants were offered a scale and polish. At each attendance, participants were asked to report any problems or adverse events and these were recorded anonymously with the experimental data.
2.5. Table 2 – Composition of the control and test dentifrices as % by weight. Control Aqua Hydrated silica Sorbitol PEG-32 Cellulose gum Sodium bicarbonate Aroma Sodium lauryl sulfate Sodium benzoate Sodium saccharin Sodium fluoride Limonene CI 77891
Test 58.38 17.50 17.50 1.39 1.20 1.00 0.79 0.75 0.50 0.26 0.22 0.01 0.50
Aqua Hydrated silica Cellulose
26.50 6.39 10.00
Cellulose gum Sodium bicarbonate Aroma
1.00 5.00 0.97
Sodium saccharin Sodium fluoride Limonene CI 77891 Glycerin Pentasodium triphosphate Sodium polyaspartate 2000–3000 g/mol Disodium phosphate Cocamidopropyl betaine Sodium phosphate Zinc citrate Bromochlorophene
0.26 0.32 0.03 0.75 39.92 3.00 2.00 1.00 1.30 1.00 0.46 0.10
Reliability of measurements
Fifteen participants had repeat measurements made of the parameters recorded by both clinical examiners (AR and AKJ), to determine the inter and intra-examiner reliability. Repeat measurements were performed by examiners at least one hour after the first measurements and were undertaken blind to them. Intra-examiner repeatability assessment of AR gave R = 0.65, 0.95 and 0.85 for mean stain, mean calculus index and mean MGI respectively. For AKJ the equivalent values were R = 0.79, 0.94 and 0.90. These data indicate that both examiners show substantial to excellent repeatability of measurements. For inter-examiner reproducibility R = 0.60, 0.92 and 0.71 for mean stain, mean calculus index and mean MGI respectively, indicating moderate to excellent reproducibility of measurements.
2.6.
Data analysis
Data were entered onto Excel spreadsheets using the participant code to maintain the anonymity of the participants, and then transferred to the PASW18.0.3 (formerly SPSS, IBM Corporation, Somers, NY 10589, USA) statistical package for
316
journal of dentistry 41 (2013) 313–320
detailed analysis. The participants’ age, gender and domicile postcodes were obtained from hospital records. The GeoConvert tool (http://geoconvert.mimas.ac.uk/) was used to link participants’ postcodes with their Index of Multiple Deprivation 2007 (IMD) score21 as a measure of socioeconomic status. IMD uses 38 indicators to assess access to resources for 32 482 areas in England. The culmulative impact of income, health, education and other domains can be used for Government planning but has been shown to relate to health in both rural and urban areas.22 Histograms confirmed the data were normally distributed before further analysis. All statistical tests of hypotheses were two-sided, and employed a level of significance of a = 0.05 at 95% confidence level. The intra- and inter-examiner data were analysed by calculation of the intraclass correlation coefficient (ICCC). Stain, calculus score and MGI scores were compared at 3 and 6 months, using a generalised linear model analysis with Wald’s Chi-square test for hypothesis testing using the follow-up data as the dependent variable, the dentifrice as the fixed factor and the baseline scores as the covariate. These analyses were undertaken on an intention to treat basis using both principled (multiple imputation) and unprincipled (last observation carried forward) methods for imputating missing data.
2.7.
Multiple imputation
Little’s test was used to demonstrate that the missing data were missing completely at random (MCAR p = 0.8) and thus amenable to imputation.23 For principled imputation a regression model was utilised using the complete baseline covariate dataset as predictor and longitudinal data as both predictor and to be imputed where missing. A constraint was applied that all imputed data must be 0 as the clinical indices cannot be negative. The three outcome measures were imputed independently. 20 data sets were imputed for each outcome as preliminary analyses showed that with this number of iterations the means tended to a constant value (data not shown). Means and standard error of the means were obtained of the pooled original and imputed data sets.
Fig. 1 – Flow of participants through the trial.
3.1.
As shown in Table 3, the two cohorts were similar with respect to demographic characteristics.
3.2. 2.8.
Adverse events
Last observation carried forward
A widely adopted unprincipled imputation method is last observation carried forward (LOCF) and this was done by substituting zero for missing stain and calculus indices and baseline values for gingival index.
3.
Demographic data
At 3 month review one participant in each of the test (1/38) and control (1/43) groups reported that ulcers were lasting longer than usual. No pathology was visible on examination. Some participants (18/92) commented upon the ‘dull’ or ‘bland’ taste of both of the dentifrices.
Results
Ninety-two participants were recruited to the study with 46 in each of control and test groups. At 3 months, 11 participants (3 control and 8 test) had withdrawn from the study leaving 81 participants available for data analysis. At the end of the study a further 3 control participants had withdrawn leaving 78 participants available for data collection (40 control and 38 test) as detailed in Fig. 1.
Table 3 – Summary demographic descriptive data showing gender, age and Index of Multiple Deprivation 2007 score21 for the control and test groups.
Gender Age IMD 2007
Control
Test
19 male, 27 female Range 28–64 years, x¯ = 43.3 x¯ = 18.6, SD = 14.9
20 male, 26 female Range 21–59 years, x¯ = 37.5 x¯ = 18.3, SD = 13.3
317
journal of dentistry 41 (2013) 313–320
Table 4 – Mean values of clinical measurements at baseline and imputed datasets at 3 and 6 months with 1 standard error of the mean in parentheses. Extrinsic stain
Baseline 3 month 6 month
3.3.
Calculus index
Test
Control
Test
Control
Test
Control
4.88 (0.53) 3.04 (0.41) 5.11 (0.55)
5.43 (0.52) 4.17 (0.40) 6.14 (0.55)
0.64 (0.044) 0.33 (0.034) 0.40 (0.038)
0.71 (0.058) 0.38 (0.031) 0.44 (0.038)
0.62 (0.059) 0.33 (0.023) 0.32 (0.022)
0.65 (0.052) 0.33 (0.023) 0.27 (0.021)
Baseline comparisons
At baseline, the data for all parameters were normally distributed and the two groups were similar (Table 4 and Fig. 2a–c).
3.4.
Mean gingival index
Hypothesis testing from imputed data sets
Means and standard errors of the means for both baseline and imputed datasets are presented in Table 4 and Fig. 2a–c. All
three clinical measurements were markedly reduced 3 months after professional cleaning at baseline. Extrinsic staining and calculus showed an increase between 3 and 6 months whereas the mean gingival index maintained its reduced level. Table 5 summarises p values from hypothesis testing on the imputed datasets. MI and LOCF show a statistically significantly greater effect of the test dentifrice on preventing dental staining at 3 months.
4.
Discussion
4.1.
Relative efficacy of dentifrices
The aim of this double-blind randomised controlled trial was to test the efficacy of a polyaspartate-containing dentifrice in controlling deposition of dental stain for 3 and 6 months. The novel dentifrice effectively reduced development of stain compared to a control fluoridated toothpaste at 3 months and the difference between the dentifrices was the same magnitude but not significant at 6 months. The dentifrices were similarly effective with respect to their ability to control calculus deposition and gingival inflammation. The Relative Dentine Abrasivity scores for the test and control were 48.1 and 38.5 respectively, which while not matched, indicate that the dentifrices both have relatively low abrasivity by comparison with proprietary ‘whitening’ toothpastes where scores over 100 are commonplace.24 The apparent diminution of effect at 6 months appears to be caused by three features of the study design, whose effects on the analysis are culmulative: (a) At 3 months the test group had adjusted mean stain scores around 20% lower than the controls group and 5% lower adjusted mean calculus scores. These effects are substantially smaller than those predicted when the sample size was estimated before the trial and show that a larger
Fig. 2 – (a) Extrinsic stain scores (shown as MI mean W 1 S.E.M.) in test and control dentifrices at baseline, 3 and 6 month review (* indicates a statistically significant difference between the dentifrices with p < 0.05 with baseline scores as a covariate). (b) Calculus scores (shown as MI mean W 1 S.E.M.) in test and control dentifrices at baseline, 3 and 6 month review. (c) Mean Gingival Index (shown as MI mean W 1 S.E.M.) in test and control dentifrices at baseline, 3 and 6 month review.
Table 5 – Summary of hypothesis testing on imputed datasets using Wald’s Chi-squared test. Extrinsic stain
Calculus index
Mean gingival index
Multiple imputation 3 month p = 0.04 p = 0.19 6 month
p = 0.27 p = 0.50
p = 0.90 p = 0.08
Last observation carried forward p = 0.01 3 month 6 month p = 0.09
p = 0.27 p = 0.54
p = 0.90 p = 0.35
318
journal of dentistry 41 (2013) 313–320
sample is required to reveal significant differences. Previous studies of the relative effectiveness of test dentifrices in either preventing staining de novo or removing pre-existing staining show substantial differences. For removal of forced stain, Moran et al.,25 found negligible benefit from a test dentifrice; whereas Isaacs et al.,26 found that 3 dentifrices inhibited all stain development for 3 months. (b) The trial suffered from substantial loss to follow-up with a resultant loss of power during ITT analysis. It is recommended for future trials that loss to follow-up is minimised by making rigorous attempts to maintain contact with participants. (c) There is a greater variance at 6 months due to a combination of a greater proportion of imputed data and an inevitable increase in variance of clinical data. Dental staining is thought to arise through a combination of three processes. First, by complexation of metal ions, such as those of copper, iron and tin into the pellicle. This may be brought about by agents within mouthwashes or by plaque microorganisms.27 Polyaspartate has been shown to reduce bacterial adhesion to hydroxyapatite in vitro.15 As yet the mechanism for this is unknown but it is likely that potent binding between aspartate and the mineral phase restricts bacterial binding to the mineral15 by competing with extracellular adhesion proteins. It might be expected that such competition may lead to a quantitative reduction in plaque on tooth surfaces. However this was not seen in a short in vivo experiment28 although a qualitative change in the plaque cannot be excluded. Fluoropolymers have been shown to reduce binding of mixed bacterial cultures to hydroxyapatite29 but showed only a modest stain inhibiting effect suggesting that other modes of staining are clinically significant. Second, clinically-visible dental stain contains only traces of reduced metals and is primarily a layer of highly calcified organic material.30 Polyaspartate is able to sequester calcium ions and inhibit precipitation of calcium phosphate-containing salts at physiological pH31 and thus it is likely that the reduction in staining seen in this trial may be due to the test dentifrice’s ability to reduce mineralisation. Although sodium polyasparate is known to be able to dissolve hydroxyapatite and brushite32 this effect is greatest at a pH significantly more alkaline than that found in the oral cavity and this will limit the potentially destructive action, as will the short duration of exposure to the dentifrice. Third, many coloured vegetable foodstuffs contain polyphenols which can bind to hydroxyapatite via proline rich salivary proteins.33 Polyaspartates are likely to compete with proline-rich proteins at the tooth surface thereby limiting adherence of polyphenols. This model of extrinsic staining was explored by Shellis et al.,34 who showed that polyphosphates could desorb protein and polyphenolic stains from hydroxyapatite in vitro. Their data indicated that a reduction in staining was maximal when the experimental agent could compete electrostatically with protein at the tooth surface and chelate the calcium salts of chromogens. Both polyphosphate and polyaspartate fulfil these criteria.
In addition to the modes of adhesion outlined above, adherent proteins and blocking agents are subject to degradation by host and bacterial enzymes. Thus the period for which an agent is resident on the tooth surface and able to resist adhesion of dental stain may be quite brief.35 One may therefore expect that dental staining develops progressively and that its rate of clinical appearance is delayed by the polyaspartate-containing dentifrice resulting in its reestablishment at the 6 month examination. The Shaw and Murray Stain Index18 is a technically simple procedure relying on the subjective transfer of the appearance of dental stain onto a grid by the examiner. Other studies have developed more objective assays utilising spectrophotometry, image analysis36 or quantitative light-induced fluorescence (QLF).37 These techniques do not remove subjectivity from assessment of staining as demarcation of areas of interest and positioning of the test teeth is highly dependent on the operator. Nonetheless they have achieved excellent intra-operator repeatability in vitro. However, currently the key barrier to their use is the requirement for accurate alignment of the apparatus in relation to the test teeth. In this clinical trial the intra- and inter-operator repeatability and reproducibility for the Shaw and Murray Stain Index were sufficient.
4.2.
Interpretation of imputation
Ideally all participants in a clinical trial would receive all interventions and be followed up at intervals prescribed by the protocol. However it is inevitable that some will be lost, perhaps as a consequence of the intervention.38 Various statistical methods exist for imputation of missing data but as it is impossible to prove that data is MCAR, none of these is completely satisfactory.38 It has therefore become commonplace for more than one method of imputation to be used, ideally including MI,39 and a conservative judgement made based on the statistical findings. Whilst it is argued that MI is dependent on the quality of model, unprincipled methods all tend to reduce the participant variation in the data.38 It is often believed that LOCF is conservative but in this study LOCF will tend to overestimate the efficacy of dentifrices by carrying forward the zero indices resulting from professional cleaning. Nonetheless the imputation methods converge in finding a statistically significant effect of the test dentifrice on dental staining at 3 months.
4.3.
Outcomes
This trial has shown, in agreement with previous studies, that a dentifrice containing agents that competitively bind to hydroxyapatite will reduce development of dental stain and calculus. The experimental polyaspartate dentifrice had a significant effect on stain formation, compared to the control dentifrice in the 3 months after dental prophylaxis. The experimental dentifrice appears effective at retarding the development of stain on recently cleaned tooth surfaces. However, the efficacy in controlling stain at 6 months should be interpreted with caution due to the differences being clinically small at this interval.
journal of dentistry 41 (2013) 313–320
Conflict of interest statement Professor Rawlinson and Dr. Jowett had full access to all the data in this study and take responsibility for the integrity of the data and the accuracy of the data analysis. Neither of these authors has a commercial or financial interest with Boots UK Ltd., Nottingham, UK. Mr. Marlow of Boots UK Ltd. contributed his experience only to the study design and did not participate in the clinical study. Boots UK Ltd. supplied the research materials and partially funded the clinical costs. This research was supported by research grants from the School of Clinical Dentistry, University of Sheffield and University of Sheffield and Boots National Centre for Applied Research in Oral Health.
Acknowledgements We are grateful to Mrs. S. Witham and Mrs. C. Hagglington for their clinical assistance, Mrs. P. Baker and Mrs. S. Wan for coordinating the research clinics, Dr. D. Robinson, Statistical Services Unit, The University of Sheffield for his advice on study design and Professor P.G. Robinsion, Academic Unit of Dental Public Health, The University of Sheffield for his advice on intention to treat analysis.
references
1. Kinane DF. Causation and pathogenesis of periodontal disease. Periodontology 2000 2001;25:8–20. 2. National Audit Office. Reforming NHS Dentistry: Ensuring effective management of risks. Report by the Comptroller and Auditor General; 25 November 2004. 3. Watts A, Addy M. Tooth discolouration and staining: a review of the literature. British Dental Journal 2001;190: 309–16. 4. Walsh TF, Rawlinson A, Wildgoose D, Marlow I, Haywood A, Ward JM. Clinical Evaluation of the stain removing ability of a whitening dentifrice and stain removing system. Journal of Dentistry 2005;33:413–8. 5. Xie P, Lu J, Wan H, Hao Y. Effect of toothpaste containing dlimonene on natural extrinsic smoking stain: a 4-week clinical trial. American Journal of Dentistry 2010;23:196–200. 6. Prasad KV, Sreenivasan PK, Rajesh G, Ramya K, Rao CB, DeVizio W. The efficacy of dentifrices on extrinsic tooth stains among community dwelling adults in India – a randomised controlled trial. Community Dental Health 2011;28:201–5. 7. Joiner A. Whitening toothpastes: a review of the literature. Journal of Dentistry 2010;38(Suppl. 2):e17–24. [Epub 2010 May 24]. 8. Tere´zhalmy G, Chaves E, Bsoul S, Baker R, He T. Clinical evaluation of the stain removal efficacy of a novel stannous fluoride and sodium hexametaphosphate dentifrice. American Journal of Dentistry 2007;20:53–8. 9. Schiff T, Saletta L, Baker RA, He T, Winston JL. Anticalculus efficacy and safety of a stabilized stannous fluoride/sodium hexametaphosphate dentifrice. Compendium of Continuing Education Dentistry 2005;26(9 Suppl. 1):29–34.
319
10. Winston JL, Fiedler SK, Schiff T, Baker R. An anticalculus dentifrice with sodium hexametaphosphate and stannous fluoride: a six-month study of efficacy. Journal of Contemporary Dental Practice 2007;8:1–8. 11. Schiff T. Anticalculus effect of a cetylpyridinium chloride/ zinc gluconate mucoadhesive gel: results of a randomized, double-blind, controlled clinical trial. Journal of Clinical Dentistry 2007;18:79–81. 12. Llena C, Forner L, Vento C. Anticalculus efficacy of a new dentifrice. Quintessence International 2009;40:497–501. 13. Fairbrother KJ, Heasman PA. Anticalculus agents. Journal of Clinical Periodontology 2000;27:285–301. 14. Sowinski JA, Guido WB, Delores MP, Petrone ME, De Vizio W, Volpe AR. A clinical study to assess the anticalculus efficacy of a new dentifrice containing a special grade of silica (Colgate Total Plus Whitening Toothpaste): a clinical trial on adults. Journal of Clinical Dentistry 2002;13:65–8. 15. Guan YH, Lath DL, deGraaf T, Lilley TH, Brook AH. Moderation of oral bacterial adhesion on saliva coated hydroxyapatite by polyaspartate. Journal of Applied Microbiology 2003;94:456–61. 16. Sikes CS. Inhibition of mineral deposition by phosphorylated and related polyanionic peptides. US Patent 5,051,401; 1991. 17. Pocock SJ. Stratified randomization. Clinical trials: a practical approach. Chichester: Wiley; 1983. 80–89. 18. Shaw L, Murray JJ. A new index for measuring stain in clinical trials. Community Dentistry and Oral Epidemiology 1977;5:116–20. 19. Volpe AR, Manhold JH, Hazen SP. In vivo calculus assessment. Part I. A method and its examiner reproducibility. Journal of Periodontology 1965;36:292–8. 20. Lobene RR, Weatherford T, Ross NM, Lamm RA, Menaker L. A modified gingival index for use in clinical trials. Clinical Preventive Dentistry 1968;8:3–6. 21. Noble M, McLennan D, Wilkinson K, Whitworth A, Barnes H, Dibben C. The English Indices of Deprivation 2007, London: communities and local government. Available at: http:// www.communities.gov.uk/documents/communities/pdf/ 733520.pdf; 2008 [accessed 9.8.11]. 22. Jordan H, Roderick P, Martin D. The index of multiple deprivation 2000 and accessibility effects on health. Journal of Epidemiology and Community Health 2000;58:250–7. 23. Kenwood MG, Carpenter J. Multiple imputation: current perspectives. Statistical Methods in Medical Research 2007;16:199–218. 24. Drisko C. Oral hygiene and periodontal considerations in preventing and managing dentine hypersensitivity. International Dental Journal 2007;57:399–410. 25. Moran J, Claydon MCA, Addy M, Newcombe R. Clinical studies to determine the effectiveness of a whitening toothpaste at reducing stain (using a forced stain model). International Journal of Dental Hygiene 2005;3:25–30. 26. Isaacs RL, Bartizek RD, Owens TS, Walters PA, Gerlach RW. Maintenance of tooth color after prophylaxis: comparison of three dentifrices. Journal of Clinical Dentistry 2001;12:51–5. 27. Addy M, Moran J. Mechanisms of stain formation on teeth, in particular associated with metal ions and antiseptics. Advances in Dental Research 1995;9:450–6. 28. Claydon NC, Addy M, Newcombe R, Moran J. The prevention of plaque re-growth by toothpastes and solutions containing block copolymers with and without polypeptide. Journal Of Clinical Periodontology 2005;32:545–8. 29. Churchley D, Rees GD, Barbu E, Nevell TG, Tsibouklis J. Fluoropolymers as low-surface-energy tooth coatings for oral care. International Journal of Pharmaceutics 2008;352: 44–9. 30. Tantbirojn D, Douglas WH, Ko CC, McSwiggen PL. Spatial chemical analysis of dental stain using wavelength
320
31. 32.
33.
34.
35.
journal of dentistry 41 (2013) 313–320
dispersive spectrometry. European Journal of Oral Science 1998;106:971–6. Sikes CS, Wheeler AP. Regulators of biomineralization. Chemtech 1988;18:620–6. Littlejohn F, Eduardo Sa´ez A, Grant SS. Use of sodium polyaspartate for the removal of hydroxyapatite/brushite deposits from stainless steel tubing. Industrial and Engineering Chemical Research 1998;37:2691–700. Proctor GB, Pramanik R, Carpenter GH, Rees GD. Salivary proteins interact with dietary constituents to modulate tooth staining. Journal of Dental Research 2005;84:73–8. Shellis RP, Addy M, Rees GD. In vitro studies on the effect of sodium tripolyphosphate on the interactions of stain and salivary protein with hydroxyapatite. Journal of Dentistry 2005;33:313–24. Jin Y, Yip H-K. Supragingival calculus: formation and control. Critical Reviews in Oral Biology & Medicine 2002;13:426–41.
36. Lath DL, Smith RN, Guan YH, Karmo M, Brook AH. Measurement of stain on extracted teeth using spectrophotometry and digital image analysis. International Journal of Dental Hygiene 2007;5:174–9. 37. Adeyemi AA, Jarad FD, de Josselin de Jong E, Pender N, Higham SM. The evaluation of a novel method comparing quantitative light-induced fluorescence (QLF) with spectrophotometry to assess staining and bleaching of teeth. Clinical Oral Investigation 2010;14: 19–25. 38. Lachin JM. Statistical considerations in the intent-to-treat principle. Controlled Clinic Trials 2000;21:167–89. 39. Blough DK, Ramsey S, Sullivan SD, Yusen R. The impact of using different imputation methods for missing quality of life scores on the estimation of the cost-effectiveness of lung-volume-reduction surgery. Health Economics 2009;18:91–101.