The effectiveness of 10% chlorhexidine varnishtreatment on dental caries incidence in adults withdry mouth

David W. Banting', Athena Papas\ D. Christopher Clark\ Howard M. Proskin^ Miklos Schultz^and Ross Perry*'School of DentisU7,1'niversiiy of Western Ontario, London, ON, Canada. -School of Dental Medicine, Tuft's University, Boston, MA,USA. 'Faculty of Dentistry, University of British Columbia, Vancouver, BC, Canada. ^Howard M. Pniskin and Associates, Inc., Rochester.N'l, USA. 'SciAn Services, Inc., Toronto, ON, Canada. "BIOS, Inc.. Toronto, ON, Canada.

Abstract

Objectives: This study compared a 10% chlorhexidine varnish treatment with placebo and sham treatments

for preventing dental caries in adult patients with xerostomia (dry mouth). Design: The study was a multi-

centred, randomized, parallel group, double blind, placebo-controlled clinical trial. Setting: All

examinations and procedures were performed at Tuft's University, Boston, MA, the University of British

Columbia, Vancouver, BC or the University of Western Ontario, London, ON. Subjects: Subjects were

adults with recent or current dental caries experience, high salivary levels of cariogenic microorganisms

and low salivary flow rates. Results: 236 subjects completed at least one post-treatment examination.

There were 697 new carious lesions diagnosed, 446 (64%) located on coronal surfaces and 251 (36%)

located on root surfaces. The mean attack rate was 0.23 surfaces /100 surfaces at risk. A treatment difference

observed between the Active and Placebo groups was statistically significant for root caries increment

(p=.O2) and total caries increment (p=.O3). A treatment difference observed between the Active and Sham

groups was not statistically significant for coronal, root or total caries increment. Analysis of variance of

treatment group differences was performed using mutans streptococci counts, salivary flow rates, age,

sex, caries prevalence, medications, time to first event and early withdrawal as co-variables. These factors

did not meaningfully alter the findings. Conclusions: The difference between the 10% chlorhexidine

vamish and placebo treatments is considered to be highly clinically significant for root caries increment

(41% reduction) and for total caries increment (25% reduction) but only for coronal caries increment

(14%).

Key words: Dental earies, prevention, root earies, ehlorhexidine, adults, dry mouth

Background

Dental caries development involves themetabolism of carbohydrate by oral bacteria in theoral biofilm leading to the production of acidswhich diffuse into the mineralized tooth tissues.These acids cause the minerals to dissolve, aprocess known as demineralization. This processtakes place within a deposit (plaque) covering thesurface of a tooth at any given site on the toothwhich is open (exposed) to the oral cavity

environment in such a way that, over time, theoutcome of this process is a disturbance of theequilibrium between the mineralized tissue(enamel, dentin, cementum) and its ecosystem'. Ifremineralization does not take place at asufficiently compensatory rate, a surface defectresults.

Numerous clinical studies ha\e implicatedmutans streptococci (MS) and lactobacilli (LB) in

Volume 17, No. 2

OS i i .n i . lW ll.nmn,.. MIUMM l',.p,.s. D(Lirk. Itowmil M. Pri.skin. Mikli.s

the developiiKtit ol coronal caries. Streptococci

.lie Icli to be instiumenial in the initiation of the

lesion atui l.ictobacilli in its progression''. These

two spci. k's ol li.tcU'iia have also been associated

with Ihe ilc\ i.lo|tmi'nt of root caties"*''. Although

IIK- loc.iiion on the tooth vv here the lesion develops

mav v.ity, the aetiologv and palhogenesis of

coional aiul tool oaiies are similar'"" and may bc

niaoifesMiions of the same dtsease process ' .

W hether a carimis lesion dcv elops on tlie coronal

ot toot surlacc of a tooth is a function of w here

ilcni.tl plaque is deposited. Plaque accumulates in

the proximity of the crest of the gingival margin.

If the gingival margin approximates the crow n of

the tooth, caries occurs on the enamel, other

ciMiditioiis beitig favourable. If the gingival margin

approximates the \oo\h root (following recession

ol the ginglva and/or loss of attachment), root

cartes occuts. other conditions beitig favourable.

\ positive correlation between the incidence ol

coronal and root canes has been established. The

relativ e odds of developing root caries are 4.3 times

htghet in indiv iduals who had previouslv

developed coronal caries compared to those who

had not".

. \cidogenic, aciduric bacteria (MS and LB)

present in dental plaque are considered to be the

pr imary risk factors in the ini t iat ion and

pio'jrcssion ol dental cartes. Lower bacterial counts

are generallv associated with decreased caries

incidence. 'Ihcv are iiecessarv. but not sullicient

lor dental caries to occur. The presence of a

susceptible tooth surlaee and a substrate suitable

for bacterial metabolism also need to be present at

the same time and for a sufficiently long period of

time. Factors that determine tooth susceptibility

and the presence of carbohvdrate substrate are

tcferred to as risk indicators. Risk indicators for

denial caries include, but are not necessarily limitedto, previous (coronal) caries experience'^'Mactors

causing reduced salivary How, lifestyle factors suchas diet, otal care and exposure to lluoride.

Chlothexidine has been shown to significantlyreduce the levels of mutans streptococci in adults"'". The elfectiveness ol chlorhexidine is attributablein large part to its substantivity (i.e. the ability tobe retained in the oral environment for prolongedperiods at an effective concentration) as well as itspotency ' -^

Thts study was designed to determine whether achlorhexidine varnish would reduce the incidenceof dental caries in adults with reduced salivary fiowdue to medication and who ate considered to be athigh risk of the disease.

Methods and Materials

The study design was a multi-centred, randomized,parallel group, double-blind, placebo-controlledclinical trial. A third (sham) treatment was includedto mimic the bitter taste of the active ingredient.The primal^ putpose of the study was to determinewhether chlorhexidine varnish treatment wassuperior at controlling and/or preventing dentalcaries in adult patients with xerostomia (dry mouth)caused by medication compared to either theplacebo or the sham treatment.

Subjects were recruited into the study using strictinclusion and exclusion criteria (Eigure 1). Everyeffort was made lo recruit subjects who wereconsidered to be at high risk of dental caries attack.Recent or current dental caries experience, highsaliv ary levels of cariogenic micro-organisms andlow saliv ary fiow rates were considered importantpredisposing factors.

All examinations and procedures wereperformed at one of three clinics located in the

Figure 1. Studv inclusion/exclusion criteria

Inclusion• 45-7.^ \ears nl age• male or non-pregnant female• informed consent• 10 or more teeth• stable medication reyinicn tor 3+ months• unstiniulated saliv arv llovv rate < 0.5 niL/minute• active, untreated caries or at least 2 filled tooth surfaces in past 2 years• mutans streptococci count > 2.'iO.()O() ("HI 7mL saltva on two separate plating tests• if untreated caries present, only one muiaiis streptococci count > 25().(J(M) CFU/ mL saliva required

Exclusion• poor oral health (soft tissue lesions, advanced periodontal disease, candidiasis. salivary gland dysfunction)• ec/ema. asthma, allergies to products• recent professional topical fluoride application• more than 10 surfaces requiring restnratuc Ireatment

£) The f i 2(MH)Gerodontology

l i e x i i l i n e \ ; i m i s h t i c i i t n i c i i t o n i k ' i i U i l c . i i i e s i n c i i k i u c I n i i d u l l s w i l h d r y m n i i l h

schools at TuIVs University, Boston. MA,the University of British Columbia. Vancouver. BCand the University of Western Ontario. London.ON. Altogether, there were thirteen clinic visitsscheduled over approximate ly 15 months(Figure 2). Dcnliil ciirics. uiistimulated andstimulated saliviiry flow rates and soil tissueexaminations were evaluated bimonthly lor nni:year. The s t imulated saliva was used formicrobiological testing. Bitewing radiographs weretaken before the treatment was applied and at thesubject's last examination.

If a subject experienced two or more carioussurfaces in any one twn-month period or more thanfour carious surfaces during the entire study, he/she w as withdrawn from the sttidy. the lesions wererestored and aggressi\e topical lluoride therapyw as provided. For subjects remaining in the study.all new carious lesions that developed during thestudy were restored within two weeks followingthe examination at which they were obserxed.

Subjects were requested not to have topicalfluoride applied to their teeth hy a dentist or other

health piokssional throtighout the observationperioil, StihjcLls were also asked to report anyunusual symptoms or occurrences to theinxestigaloi inmiedialely. Sei lous and tmexpectedadverse experiences were extensively investigatedas to symptoms, signs, time of onset, duration,therapy outcome and assessment of any possiblerelationship to the tesi medications.

Subjects were stratified by age. number of teethand MS levels and then randomized to one of threestudy groups. Treatment consisted ol" a once-weekly, two-siage application by a dental hygienistover lour consecutive weeks and then a single re-application after six months following apredetermined protocol. The composition of theActive. Placebo and Sham treatments are presentedin Figure 3. Packaging was standardized andpro\ ided by a commercial contractor. Dispensingrecords were carefully maintained to determinedrug dosing.

Caries examinations were eonducted using thePitts and Fyfle-^ adaptation of the WHO caries

Figure 2. StuJs schedule

Preparation (2 months) \isits 1-3screenini:. history, salixary f\o\\ rate, mieriihiolniiy. mediealimis. restorations

Treatment (4 weeks) visits 4-7four treatment applicatinns u iih a single- re-application alter d months

Observation (12 months) visits 8-13bimonlhh examinations• eoronal and root caries• soil tissue condition• unstimulated sali\ary How late/microliioloyyAll visitsmedicatii>ns. ad\ersc e\ents. oral hyyiene instruetion. restoialioiis as required

Figure 3. Composition ot treatments hy group and stage

Stage 1(imL)

Stage 2(1 mL)

Acti\e

ChlorhexidineAeelateB.P. 10'.(w/\)

Ben/oin SumatraLi.S.P

2()'/f (w/v)

Alcohol.Dehydrated U.S.Pto volume

Placebo

Ben/oin SumatraU.S.P20'. (w/\)

Alcohol,Dehydrated U.S.Pto Nolume

Polvuiethane 2^''^ (w/w)

l-.thvl Acclale 2 2 ' . (w/wlAcetone 4')S i\\/u )

VoluiiK- 17. No. 2

Sbam

Quininehydrochloride0.2'; (w/v)

Hydroxypropy 1celluloseI'/r (u/ \)

Alcohol.Dehydrated U.S.P,water to volumeFood colouring to match

70 Dnvid \V Biiiitinn. Athena Papas. D. Christopher Clark. Howard M. Proskin. Miklc

terminology that was expanded to include root performed to test the study hypotheses. Acaries (Figure 4). lixaminer calibration wasperformed ptior to the study and examinerreliability was assessed at that time and again after(i montlis lhrec examiners were used in this study,one .It each site. Examiner reliability was measuredusing the intra-class correlation coenicient".

The prituary efficacy parameter lor this study isthe caries increment score expressed as new(primar\ or secondary) carious tooth surfaces persubject. Caries increment was determined using apairwise tnatrix that contrasts the status of a toothsurfaee at the baseline examination (visit 7) withIts status at each of the subsequent visits over thenext 12 months and takes into account incrementalpairtngs. illogical pairings and pairings that resultin no effect. Each central, lateral and cuspid toothis deemed to have four coronal and four rootsurfaces u hereas each posterior tooth is deemedto have five coronal and four root surfaces.

Statistical analyses were performed for cariesincrements occurring on both coronal and rootsurfaces. For subjects who withdrew from the studyvoluntarily or because they were considered to bea treatment failure (developed more than fourcarious surfaces during the study), the cariesincrement score was determined at their lastexamination (last observation carried forward). Allcomparative statistical tests of significance are two-sided and performed at the 0.05 alpha level ofsignificance. The caries increment scores wereranked using the SAS Proc Rank procedure^".Efficacy analysis was performed on an intent-to-treat basis. Analysis of variance of the rank-transformed increment scores was performed usingthe SAS General Linear Models procedure.Analysis of covariance was used to determine theinfluence of age. number of teeth and MS level ontreatment effect. Prior to pooling the data over sites,a site by treatment interaction was investigated. Apriori, pairwise comparisons using t-tests were

Figure 4. Diagnostic criteria for coronal and root caries*

difference in efficacy of at least 15% is consideredto be clinically significant.

Results

Five hundred and twenty-one subjects werescreened and 240 (46%) of these subjects wereenrolled in the study. Four subjects dropped out ofthe study before the treatment visits werecompleted leaving 236 subjects who attended atleast one follow-up examination visit. One hundredand eighty (76%) subjects completed the entire 13-visit protocol. Subject withdrawal and completionexperience were similar at all three sites. Treatmentfailure (more than 4 decayed surfaces) accountedfor 21 % of withdrawals from the study. Caries datawere not available for two subjects.

The characteristics of the study population arepresented in Table 1. The mean age of the studysubjects was 58.7 years with a standard deviationof 11.4 years. The subjects at the UWO site were,on average, slightly younger. There were morefemales (59%) than males (41%) with no sitedifferences. Most subjects were Caucasian (92%).Blacks represented 10% of the subjects at the Tuftssite and Asians represented 11 % of the subjects atthe UBC site. About 26% of the subjects reportedthat they were smokers, 23% wore dentures, 23%used mouthwash regularly and 84% used afiuoridated dentifrice. No large site differenceswere observed for reported smoking, use ofmouthwash and use of a fluoride dentifrice. Themajority of subjects at the UBC site did not usefluoridated water (Vancouver's water supply is notfluoridated) whereas at the two other sites, themajority of subjects used fluoridated water.

Almost all of the study subjects were takingmultiple medications on a long term basis. Themost common prescription drugs consumed wereanti-hypertensives and anti-depressants accountingfor more than 65% of all medications reported. The

CodeDO/T^O

Dl /R l

D2/R2

D3/R3f)4

CategoryCoronal soundRoot soundCoronal initial

Root initialCoronal enamel

Root dentineCoronal dentinePulp involvement

Criteriano evidence of cariesno evidence of cariesno clinically detectable loss of substance-staining in pits and fissures, white spots on smooth surfacesloss of tooth surface integrity (cavitation)loss of tooth surface integrity (cavitation). lesionconfined to enamelcavitation involves loss of dentinecavitation has reached dentinenot used

* Pitts and Fyffe"D= primary cariesR= recurrent (secondary) caries

& The Gcrodontology Association 2000Gerodontology

c varnish treulnii'm mi denial caries MiLideni.L" m adults with dry mouth 71

mean unslimulated saliva flow rale was O.OX mL/min. Subjtxis at the Tufts site had lower salivaryflow rates (O.O.S mL/min) eompared to the othertwo sites and this difference was statisticallysignificant (Tu key-Kramer HSD test).

The mean mutans streptococci level in the salivafor the study subjects was high at 7.9.̂ x 106 CFU/mL saliva. The UBC stte had a higher meansaliviiry mutans streptococci count compared to theother sites but this difference was only statisticallysignificant compared to the UWO site (Tukey-Kramer HSD test). Slightly more than half of thesubjects (56.6%) had carious lesions requiringrestorative treatment with one in four (25.6%)requiring four or more restorations. From thestandpoint of their past and present cariesexperience, reduced salivary flow rate and highmutans streptococci counts in the saliva, the studysubjects were considered to be at high risk of dentalcaries.

Examiner reliability for dental caries diagnosiswas acceptable for caries (primary and secondary)and excellent when caries and fillings werecombined. At the initial calibration the intraclass

correlation coefficient lor the examiners was 0.65lor caries and 0.91 for canes and fillings, l^xaminerreliability determined six monllis later remainedcon.Mslent at 0.6^ for caries and 0.90 (or caries andfillings together.

The mean caries increment score in the studypopulation was 3.0 with a .standard deviation of3.2. The median caries increment was 2.0. Seventy-four per cent of the study subjects developed cariesduring the 12-month observation period, 62%developed coronal caries and 45% developed rootcaries. Sixty-seven percent of subjects developedboth coronal and root caries.

New coronal and root caries occurred throughoutthe 12 months of observation. The pattem of newcaries increments was similar for coronal and rootsurfaces with the greatest number of lesionsoccurring in the first two months of the study.Coronal caries rates decreased and root caries ratesremained relatively constant over the remainingfive bimonthly examination periods (Figure 5).There were 697 new carious lesions diagnosed, 446(64%) located on coronal surfaces and 251 (36%)located on root surfaces. The majority of the lesions

Table 1. Selected characteristics of study subjects at baseline by site

Age (yrs)Male (7f)Caucasian (7c)Residence (yrs)Fluoridated water (7r)Use fluoride dentifrice (7f)Smoker (7c)Use mouthwash 1 or 2X dailyWears one or more dentures( 7c)Unstimulated SFR (mL/min)SM counts (CFU/mL)Active caries (7c)

Tufts

61.348.388.823.779.892.119.124723.6.05

7 36x106

UBC

60.035.787.122.02.8677.131.420.020.0.10

13.7x106*

UWO

57.537.7100

22.377.980.528.622.123.4.10

3.43x106

All

58.741.191.922.756.483.925.822.522.5.08

7.93x10657

* a single extreme value (842 x l()6i was omitted from the calculations.If included, the mean SM counts for UBC would be 25.5 xlO6 CFU/mL

Figure 5. Distribution of caries increment by type and time to event

c(1)

0}

o

to

Q)

CD

o

140

120

100

80

60

40

20

0

127

2 Months 4 Months 6 Months 8 Months

Time to event

10 Months 12 Months

Volume 17, No. 2

7 ' IXiM.I \V H,miiiijj, Aihcii^i l'iip,is, I) ("hiisioplu-i Clurk, llowaril M. Pmskiii, Miklo)-

iKctiiicil iMi IIK* proximal sui laccs ( Table 2), The

tiK.m allack i>iU' lur all unni | is was 0.2.^ suiTacos/

100 stii lai.\"s al i isk.

lahlo .̂ pii.'si.'nls the i.arios inerciiK'nl scores by

l \ | ie o l \ a i u s , ticainienl gnnip and sile. Cienerally,

the U\\ () site expeiK'iKed the highest rate olearieslolloued elosely hy the UliC site. The Tiills siteexperienced draniatieall> lower earies rates for alliieatineiit iiroiips eompared to the other two sites.Despite these seeminel\ large between-sitediHerenees, there was no statistieally signifieantvariation in the tteatmeni elTeet aeross the threesites (i.e. no signilieant treatment by siteinteraetion).

The treattiient difterenee obser\ed between the.•\eti\e and Placebo groups was statistieallysignitleant tor root cai-ies inetement (p=.O2) andlor total caries inerement (p=.O3). The differeneeobsened between the Aeti\e and Sham groups wasnot statistieall) significant for eoronal, root or totalearies ineretnent. Additional analysis of varianeeof treatment gtoup differenees was performedusing mutans streptoeoeei counts, salivary flow

Table 2. Distribution of canes increment by type

Surface Coronal Root Total

BuccalDistalLingualMesialOcclusalTotal

Table 3. Me

MS

noS5

97

56

4 4 6 ( ( i4 ' . )

an iSD) canes

S27739530

251 (3h', )

increnient b\ tv

18018712415056

697

pe ol

rates, age, sex, pievalcncc of caries at pre-treatmenland at baseline, number of medications capable ofeausing dry mouth, time to first event and earlywithdrawal as co-variables. These laelors did notmeaningfully alter the statistical findings.

The treatment difference between the Active andPlacebo gioups is eonsidered to be highly clinicallysignifieant for root caries increment (41%reduction) and lor total earies increment (25%reduction) and just falls short of the cut-off forbeing eonsideted elinically significant for coronalcaries increment (14^;^).

There were 27 serious adverse events reportedin the study, 11 (147^) in the Active group and eight(10%) eaeh in the Placebo and Sham groups. Noneof the serious adverse drug events were attributedto the study medication. Twenty of these subjects(747r) completed the study. Other adverse eventsreported primarily involved irritation of the oralmueosa and a bitter taste following application ofthe treatment. All of these adverse events weretransient, lasting less than a day.

Discussion

The subjeets partieipating in this clinical trial wereselected because they were considered to be at'high risk' of dental caries. This was determinedby the presence of an etiologieal risk factor(salivary mutans streptococci levels >250,000CFU/niL) and the presence of two risk indicators:reduced salivary flow rate (<0.5 mL/min) and

Treatmentgroup

N Coronal cariesincrement

Root cariesincrement

Total cariesincrement

AcmeTuilsUBC

UWOAll

PlaceboTultsVtiC

UWOAll

ShamTults

UWOAll

29232577

3023

79

242678

0.34(0.61)2.48(2.25)2.S4(2.90)1.79(2.36)

0.70(0.92)2.52(2.13)3.31 (2.26)2O9(2.13)

0.39(0.69)2.S3 (2.76)2 46(2.37)1.83(2.34)

0.45 (0.74)0.87(1.55)1.04(1.59)0.77(1.33)

0.97(1.19)0.96(1.26)2O0(2.2S)1.30(1.70)

0.96(1.60)1.00(1.35)1.46(1.86)1.14(1.62)

0.79 (0.90)3.35 (3.08)3.88(3.69)2.56 (3.05)

1.67(1.S3)3.48(3.01)5.31 (3.89)3.39(3.32)

U35(2.(U)3.83(3.07)3.92(3.75)2.97(3.22)

p-valuesActive vs. placeboActise vs. sham

^/i reductionAcli\e vs. placebo

p=0.06p=0.79

14.4'-/

p=0.02p=0.16

4().S9;

p=0.03p=0.36

24.5':;

('.. The I AssiiLi.'jiion 21)00

Gerodontology

-hexidiiie Viirnish ireulmenl on dental cii ius jncnlLncc in iiilulis wiili dry niduih 73

previous caries expcricllcc"•*'•-^ It was notunexpected, therefore, that dental caries wouldoccur in this study population and that it wouldoccur within a relatively short period of time.Studies (if root caries in older, adult populationshave reported new coronal and root lesions withintvlatively short periods of t ime ' " " " . The attackrates observed in this study for the placebo groupare within the tange (0.31 lo 6.3/100 surfaces peryear) of those tepoited in the litetatute. The placeboeffect associated with the incteased frequency ofcaries examinations in this study may havecontributed to a cotnpaiatively lower tate.

Van Rijkom et aP^ employed meta- analysis tosummarize the caries-inhibit ing effect ofchlothexidine tieattnent, used ptedominantly asotal rinses. Most of the studies repotted involved'high risk' subjects but they were restricted toadolescent age gtoups. The mean caries-inhibitingeftect from chlorhexidine treatment was found tobe 46% {95% CI= 35%- 57%). The frequency ofapplication of chlothexidine used in the studiesrepotted ranged from every day to every 90 days.This must be contrasted with the series of fourweekly applications followed by a single re-application six months later used in this study. Aswell, additional fluoride rinsing or application wasused in all but one of the studies reported in themeta-analysis whereas fluoride rinsing wasprohibited in this study. The vehicle used to deliverthe chlorhexidine also diffeted. The effectivenessof chlorhexidine determined in this study (24.5%)was diminished by cornparison, however, thismight be explained by the difference in thefrequency of application of the chlorhexidine and/or an additive treatment effect produced by fluorideadministration^\

Various preventive agents and regimens havebeen employed to reduce coronal and root cariesincidence in adults. The use of a daily 0.05% NaFrinse has been shown to reduce root cariesincrement by 16% over three years but have noeffect on coronal caries^". Brushing with a 1100ppm sodium fluoride dentifrice daily for one yeardramatically reduced coronal caties incidence by41% and root caries incidence by 67% in a non-fluoridated community in the United States^- butless dramatic results were found in a five monthstudy using NaF and AmF/SnF2 dentifricescombined with mouthwashes^". The addition ofsoluble calcium and phosphate ions providedsuperior results compared to a conventionalfluoride dentifrice^'. A four year study contrastingthe effectiveness of semi-annual application of1.2% APF gel and daily 0.05% fluoride rinse

demoiislialed that Ihc gel treatment reduced (lieincidence of root earies by lO'/t and Ihe rinseredtiecd ihe incidence of looi caries by 7 I 7reompatcd to a eonlrol group'-. Only one other sludyhas used a varnish vchiclc^\ Hither 40% (w/w)chlorhexidine varnish or 5'/f sodiurn fluoridevarnish ttealments were applied every three monthsfor one year. A reduction in root earies incrementof 51 % was achieved with ehlorhexidine and 36%with fluoride. In a preventive program forcomrnunity-dwelling older adults, a 0.12%ehlorhexidine rinse was used weekly with afluoride varnish applied twiee yearly and thiscombination resulted in a 27% reduction in coronalearies and a 23% reduction in root earies eventscompared to a group receiving usual care from aprivate practitioner'l The percent reduction in rootcaries incidence achieved by the application of10% chlorhexidine varnish used in this studyapproximates or exceeds that found by otherstudies and the reduction in coronal eariesincidence is generally lower. In this study, the vastmajority of subjects routinely used a fluoridateddentifrice and half of the subjects consumedfluoridated water so any treatment effect observedshould be considered additive to any beneficialeffects of brushing with a fluoridated dentifrice andconsuming fluoridated water.

The influence of surfaces filled during a studyon earies increment ean be profound. Filledsurfaces may or may not have met the study criteriafor caries. Restorations are often placed on the rootsof teeth to treat abrasion or temperature sensitivityand restorations are frequently placed to repairexisting restorations on both coronal and rootsurfaces. In fact, more than half of the filledcomponent of traditional Decayed and Filled (DF)or Root Caries Index (RCI) methods of determiningcaries increment may be attributable to fillingsplaced for reasons other than decay as defined bythe research protocol-*-""^'. In this study, only new(primary or secondary) dental caries was includedin the increment and this also may account for thecomparatively lower dental caries attack ratesobserved.

An unexpected finding in this study is thecomparatively small difference in treatment effectbetween the Active group and the Sham group. TheSham group was expected to perform similariy tothe Placebo group. The results indieate that thereis a modest treatment effect associated with 0.2%quinine hydrochloride such that the treatment effectof the Active group compared to the Sham uaselinically only marginally signifieant and notstatistically significant. The ehoiee of the sham

Volume 17, No. 2

7.) n.uul W . .\thcnn I'lipas, IX ChriMophcrClark, llowurd M. Proskin. Miklos

l in this study was inado based on the tastechaiiKtctistics ol the prodticts being tested.Chlorhovidine has a bitter taste and, in an elTort topiovont siib|oct bias, a sham material, quininehydriK hloride. was suggested beeause it has a bittertaste similar to chlorhexidine. At the time, littleconsideration was given to the antimicrobialproperties of quinine hydroehloride. In retrospect,this proved unfortunate. It was determinedfollowing the study that quinine hydroehloridedoes, indeed, have antimierobial properties. Thiswould explain why differences in earies inerementbetween the Aetive and Sham groups wereeomparatively small.

Although a statistieally signifieant difference wasnot deteeted for eoronal earies, it eannot beeoneluded that a difference does not exist. The sizeof the treatment groups employed in this study didnot provide sufficient power to detect a differenceas small as 15^^. The large amount of variabilityobserved in the earies increment seores of study groupsubjects adversely affeeted the power of this study.

Serious adverse events did occur in this studybut they were not related to the study treatments.The adverse events reported were of mild intensityand short duration and they were equallydistributed among the treatment groups. Theehlorhexidine varnish used in this study wouldappear to have a wide safety margin when used asdirected.

Although it has not been widely promoted as apreventive treatment for root earies, the results ofthis study suggest that 10% ehlorhexidine varnishmay have an important role to play in themanagement of dental earies in adults with drymouth.

References

I Fejerskov O, Baelum V, Ostergaard E S. Rootcaries in Scandinavia in the 198O's and future trendsto be expected in dental caries experience in adults.AdvDetuRes 1993; 7: II.

2. Edwardsson S. Microorganisms associatedwith dental caries. In Thylstrup A, Fejerskov O,editors. Textbook of cariology. Copenhagen:Munksgaard; 1986. p. 107 - 30.

3. Loesche W J. Role of Streptococcus mutans inhuman dental decay. Microbiol Rev 1986; 50: 353-80.

4. Bowden G W H. Which bacteria are cariogenicin humans? In: Johnson NW, editor. Dental Caries:markers of high and low risk groups and individuals.Cambridge, N.J.: Cambridge University Press; 1991:266-86.

5. Brown L R, Billings R J, Kaster A G.Quantitative comparisons of potentially cariogenicmieroorganisms cultured firom noncarious and cariousroot and eoronal tooth surfaces. Infection andImmunity 1986; 51: 765-70.

6. Ravald N, Hamp S-E, Birked D. Long-termevaluation of root surface caries in periodontallytreated patients. J Clin Periodontol 1986; 13: 758-67.

7. Ellen R P. Ecological determinants of dentalroot surface caries. In: Bowen WH, Tabak LA, editors.Cariology for the nineties. New York: University ofRochester Press; 1993. p. 319-32.

8. Scheinin A, Pienihakkinen K, Tiesko J, et al.Multifactorial modeling for root caries prediction: 3-year follow-up. Community Dent Oral Epidemiol1994;22: 126-29.

9. Loesche W J, Taylor G W, Dominguez L D,et al. Factors which are associated with dental decayin the older individual. Gerodontology 1999;16: 37-46.

10. Nyvad, B. Microbial colonization of humantooth surfaces. APMIS 1993; 101(SuppI.32): 7-45.

11. Zambon J J, Kasprsak S A. The microbiologyand histopathology of human root caries. Am J Dent1995; 8: 323-328.

12. Billings R J, Banting D W. Future directionsfor root caries research. Gerodontology 1993; 10:114-119.

© The GerixJontology Association 2(HH)Gcrodontology

acxiJine varnish trcutmcnt on dental caries incidence in adulls with dry mouih 75

13. Fure S. Five-year iticidence of coronal and rootcai-ies in 60-, 70 and 80-year old Swedish individuals.Caries Res 1997; 31: 249-258.

14. Beck J D, Drake C W. Do root lesions tend todevelop in the same people who develop coronallesions? J Pub Health Dent 1997; 57: 82-88.

15. Powell L V, Leroux B G, Persson R E, et al.Factors associated with caries incidence in an elderlypopulation. Community Dent Oral Epidemiol 1998;26: 170-176.

16. Lobene R R, Kashket S. Soparkar P M, et al.The effect of cetylpyridinium chloride on humanplaque bacteria and gingivitis. Pharmacol TherDent1979; 4: 33-47.

17. Hancock E B, Newell D H. Antimicrobials inperiodontal practice. Dent Clin North Am 1994; 38:719-731.

18. Jenkins S, Addy M, Wade W, et ai Themagnitude and duration of the effects of somemouthrinse products on salivary bacterial counts. JClin Periodontol 1994; 21: 397-401.

19. Renton-Harper P, Addy M, Moran J, et al, Acomparison of chlorhexidine, cetylpyridiniumchloride, triclosan and C31G mouthrinse products forplaque inhibition. J Periodontol 1996; 67: 486-489.

20. Allen D R, Davies R, Bradshaw B, et al.Efficacy of a mouthrinse containing 0.05%cetylpyridinium chloride for the control of plaque andgingivitis: a 6 month clinical study in adults. CompendContin Educ Dent 1998; 19(2 Suppl): 20-26.

21. Marsh P D. Microbiological aspects of thechemical control of plaque and gingivitis. J Dent Res1992; 71: 1431-1438.

22. Fine D. Evaluation of antimicrobialmouthrinses and their bactericidal effectiveness. JAmDent Assoc 1994; 125; 11S-19S.

23. Eley B M. Antibacterial agents in the control ofsupragingival plaque- a review. BrDentJ 1999;186:286-296.

24. Pitts N B, Fyffe H E. The effect of varyingdiagnostic thresholds upon clinical caries data for alow prevalence group. J Dent Res 1988; 67:592-596.

25. Fleiss J. The design and analysis of clinicalexperiments. New York: John Wiley & Sons, 1986;p. 14-25.

26. SAS Institute, Cai^, NC 27513, USA

27. Burgess R C. Assessment of caries risk factorsand preventive practices. J Dent Educ 1995; 10: 962-971.

28. Ravald N, Hamp S-E. Prediction of root cariesin patients treated for advanced periodontal disease.J Clin Periodontol 1981; 8: 400-414.

29. Banting D W, Ellen R P, Fillery E D. Alongitudinal study of root caries: baseline andincidence data, y De/ir/?£"* 1985; 64: 1141-1144.

30. Leske G S, Ripa L W. Three-year root cariesincrements: an analysis of teeth and surfaces at risk.Gerodontology 1989; 8: 17-21.

31 Hand J S, Hunt R J, Beck J D. Coronal androot caries in older lowans: 36-month incidence.Gerodontics 1988; 4: 136-139.

32. Jensen M E, Kohout F. The effect of afluoridated dentifrice on root and coronal caries inan older, adult population. J Am Dent Assoc 1988;117: 829-832.

33 Schaeken M J M, Keltjens H M A M, VanDer Hoeven J S. Effects of fluoride and chlorhexidineon the microflora of dental root surfaces andprogression of root-surface caries. J Dent Res 1991;70: 150-153.

34. Jones et al. Caries incidence in patients withdementia. Gerodontology 1993; 10: 76-82.

35. Joshi A, Papas A, Giunta J. Root cariesincidence and associated risk factors in middle-agedand older adults. Gerodontology 1993; 10: 355-361.

36. Drake C W, Hunt R J, Koch G G. Eighteenmonth coronal caries incidence in North Carolinaolder adults. J Pub Health Dent 1994; 54: 24-30.

37. Lawrence H P, Hunt R J, Beck J D. Three-year root caries incidence and risk modeling in olderadults in North Carolina. J Pub Health Dent 1995;55: 69-78.

38. Van Rijkom H M, Truin G J, van't Hof M A.A meta-analysis of clinical studies on the caries-inhibiting effect of chlorhexidine treatment. J DemRes 1996; 75: 790-795.

39. Ripa L W, Leske G S, Forte F, et al. Effect ofa 0.05% neutral NaF mouthrinse on coronal and rootcaries in adults. Gerodontology 1987; 6: 131-136.

Volume 17, No. 2

W n . i n i i i i j i , A t h c n . i I ' . i p . i s . P ( " I I I I S I I ) | M H - I t ' h i i k , l l t i w j i r d M . P m s k i n , M i k l o s ;

40. Uanoc/y .1, Ni-mcs J . I llcci ol aininc tluoridc

(Anil )/sianniius riiioritic (Snl''2) loollipasti.- and

nuuithwaslK's on ik'tilal plaiiuc aceti i iui lat ion.

giiiiiiv Ills and rool-siirlaLO caries, /'/ci h'inn Dent Soc

; 87: .̂

41. l'ap;is .\, Russell 0, Singh M, ct al. Doubleblind clinical trial o\' a rctiiincrali/ing dentifrice inthe pre\enlion of caries in a radiation therapypopulation, (niotlontology \999: 16: 2-10.

42 Wallace M C, Rcticf D H, Bradely E L. The4S-nionth inerenienl ol tool caries in an urbanpopulation of older adults partieipating in a pre\ enli vedental program. J Pith Health Dent 1993; 53: 133-137.

43. Powell L ̂ , Persson R E, Kiyak H A, ci ul.Caries prevention in a eomtnunity-dwelling olderpopulation. Caries Res 1999; 33 : 333-339.

44 Depaola P F, Soparkar P M, Kent R L.Methodological issues relative to the quantificationol looi surface caries. Gerodontology 1989; 8: 4-8.

45 Walls A W G, Silver P T, Steele J G. Impactof treatment provision on the epidemiologicalleeording of root earies. luir J Oral Sei 2000; 108:3-8.

Address for cortespondence;

Dr D W BantingFaculty of Medicine & DentistrySchool of DentistryUnivetsity of Western OntarioLondon, Ontario, Canada N6A 5C!

Tel: OOI 519-661-2111 x86130Fax: 001 519-661-3885e-mail: dbanting@julian.uwo.ea

©The Assutialion 2000Gerodontology