Connective tissue graft plus resin-modified glass ionomer restorationfor the treatment of gingivalrecession associated withnon-carious cervical lesion: arandomized-controlled clinical trial

Santamaria MP, Ambrosano GMB, Casati MZ, Nociti Junior FH, Sallum AW andSallum EA. Connective tissue graft plus resin-modified glass ionomer restoration forthe treatment of gingival recession associated with non-carious cervical lesion: arandomized-controlled clinical trial. J Clin Periodontol 2009; 36: 791–798. doi:10.1111/j.1600-051X.2009.01441.x.

AbstractBackground: The aim of this clinical study was to evaluate the treatment of gingivalrecession, associated with non-carious cervical lesions by a connective tissue graft (CTG)alone, or in combination with a resin-modified glass ionomer restoration (CTG1R).Materials and Methods: Forty patients presenting Miller Class I buccal gingivalrecessions, associated with non-carious cervical lesions, were selected. The defectswere randomly assigned to receive either CTG or CTG1R. Bleeding on probing(BOP), probing depth (PD), relative gingival recession (RGR), clinical attachmentlevel (CAL) and cervical lesion height (CLH) coverage were measured at baseline and45 days, and 2, 3 and 6 months after treatment.Results: Both groups showed statistically significant gains in CAL and soft tissuecoverage. The differences between groups were not statistically significant in BOP,PD, RGR and CAL, after 6 months. The percentages of CLH covered were74.88 ! 8.66% for CTG and 70.76 ! 9.81% for CTG1R (p40.05). The estimatedroot coverage was 91.91 ! 17.76% for CTG and 88.64 ! 11.9% for CTG1R(p40.05).Conclusion: Within the limits of the present study, it can be concluded that bothprocedures provide comparable soft tissue coverage. The presence of the glass ionomerrestoration may not prevent the root coverage achieved by CTG.

Key words: cemento-enamel junction; gingivalrecession/surgery; glass ionomer cement;surgical flap; tooth abrasion.

Accepted for publication 20 May 2009

During the past few decades, the perio-dontal literature has presented a hugenumber of clinical trials aimed to eval-uate different surgical approaches forroot coverage. It has been recognizedthat buccal gingival recession, present-ing no loss of inter-proximal periodontalattachment and bone (Miller Class I andII), can be predictably covered by avariety of surgical procedures (Roccuz-

zo et al. 2002, Cairo et al. 2008). Themain outcomes of these studies were toevaluate the complete root coverage andpercentage of root coverage achieved bythe procedures. For this, the cemento-enamel junction (CEJ) was used as thereference point.

It has also been recognized that gin-gival recession is frequently associatedwith cervical wear. Sangnes & Gjermo

Mauro Pedrine Santamaria1, GlauciaMaria Bovi Ambrosano2, MarcioZaffalon Casati1, FranciscoHumberto Nociti Junior1, AntonioWilson Sallum1 and Enilson AntonioSallum1

1Department of Prosthodontics andPeriodontics, Division of Periodontics;2Departmentof Social Dentistry, Division ofBiostatistics, Piracicaba Dental School,University of Campinas – UNICAMP,Piracicaba, SP, Brazil

Conflict of interest and source offunding statement

The authors report no conflicts of interestrelated to this study.The authors would like to thank the Foun-dation for the Development of Personnel inHigher Education – CAPES for supportingDr. Mauro P. Santamaria.Dr. Enilson A. Sallum would like to thankCNPQ for support (#300817/2007-0).

J Clin Periodontol 2009; 36: 791–798 doi: 10.1111/j.1600-051X.2009.01441.x

791r 2009 John Wiley & Sons A/S

(1976) reported that gingival recessionand a wedge-shaped defect in the cervi-cal area were often seen affecting thesame tooth. In another report (Zucchelliet al. 2006), no signs of the CEJ wereobserved in about 50% of the examinedteeth showing gingival recession, due tocervical abrasion. The presence of anon-carious cervical lesion, associatedwith gingival recession, can cause someconfusion regarding the identification ofthe CEJ location, which is often mis-taken with the coronal border of thecervical lesion.

With the loss of the CEJ, caused bythe progression of the non-carious cer-vical lesion, it could be inferred that thecervical lesion simultaneously affectsparts of the root and crown of the tooth.Therefore, it may be speculated that themost coronal zone of the non-cariouscervical lesion is mainly formed by theexposed dentin of the dental anatomiccrown. This condition makes the com-plete coverage of the associated lesion(gingival recession plus non-carious cer-vical) an unpredictable goal. There is atrend towards leaving the coronal borderof the lesion still exposed after thesurgical procedures; however, even inthe presence of complete root coverage(gingival margin at the level or pre-existing CEJ) the patient can still pre-sent dentin sensitivity (DS), associatedwith the portion of the non-cariouscervical lesion that is exposed abovethe gingival margin (located in the ana-tomic crown).

Recently, it has been shown thatgingival recessions, associated withnon-carious cervical lesions, can besuccessfully treated by glass ionomerrestoration (Santamaria et al. 2007,2008, 2009) or a composite resin (Luc-chesi et al. 2007) combined with acoronally advanced flap. After the heal-ing period, part of the restoration wascovered by the soft tissue. Good aes-thetic outcome and gingival health withno signs of inflammation, such as red-ness and bleeding on probing (BOP),were observed. Even though soft tissuecoverage was obtained in these cases,the coronal zone of the restorations inthe group treated by a coronallyadvanced flap plus restoration, or thecoronal zone of the non-restored cervi-cal lesion in the group treated only by acoronally advanced flap, remaineduncovered, probably due to the factthat the crown portion of the lesioncould not be completely covered. How-ever, there is a lack of information

derived from randomized-controlledclinical trials on the ability of othersurgical procedures to treat gingivalrecession, associated with a non-cariouscervical lesion. Therefore, the aim of thepresent study was to compare the out-come of connective tissue graft (CTG)alone or in combination with a resin-modified glass ionomer restoration inthe treatment of gingival recessions asso-ciated with non-carious cervical lesions.

Materials and Methods

Before the beginning of the study, theconsent form and the protocol of thestudy were approved by the InstitutionalReview Board of the University ofCampinas (CEP-UNICAMP 104/2005).Informed consent was signed by eachsubject after a thorough explanation ofthe nature, risks and benefits of theclinical investigation and associatedprocedures.

Study population

Forty patients, 21 males and 19 females,aged 19–71 years (mean age 36.25! 22.8 years), were included. The sub-jects were selected from the group ofpatients referred for periodontal treat-ment to the Graduate Clinic of thePiracicaba Dental School, Universityof Campinas, to participate in this study.The patients were selected from Marchof 2006 to February of 2007, accordingto the following eligibility criteria:

1. Presence of one Class I Miller gingi-val recession, associated with a non-carious cervical lesion 1–2mm deepin maxillary canines or premolars.

2. Non-smokers.3. Systemically and periodontally healthy.4. No contraindication for periodontal

surgery.5. No use of medications known to

interfere with periodontal tissuehealth and healing.

6. Probing depth (PD)o3mm withoutBOP.

7. Tooth vitality, absence of restorationon cervical area and absence ofsevere occlusal interferences in thearea to be treated.

8. No previous periodontal surgery inthe area.

The patients were referred for perio-dontal treatment based on their com-plaints (DS and/or aesthetic concerns).Because a non-carious cervical lesion

may be the consequence of a multi-factorial process, including tooth struc-ture loss caused by non-bacterial acids(erosion), traumatic toothbrushing(abrasion) and occlusal loading (abfrac-tion) (Litonjua et al. 2003, Bartlett &Shah 2006), all patients were includedin a pre-treatment programme in orderto eliminate the possible aetiologic fac-tors related to a non-carious cervicallesion and gingival recession. Oralhygiene instructions with a non-trau-matic brushing technique and a softtoothbrush were given. Patients werealso encouraged to avoid excessive con-sumption of acidic beverages or acidicfoods. When necessary, selective grind-ing was performed to remove occlusalinterferences on the teeth included in thestudy. Scaling, root planing and crownpolishing were performed as necessary.

Clinical assessments

After this initial therapy, the followingparameters were recorded: (1) full-mouth visible plaque index (FMPI)(Ainamo & Bay 1975) and presence orabsence of visible plaque accumulationat the site included in the study [plaqueindex (PI)]; (2) full-mouth sulcus bleed-ing index (FMBI) (Muhlemann & Son,1971) and presence or absence of BOPat the site included in the study (BOP);(3) PD, assessed as the distance fromthe gingival margin to the apical endof the gingival sulcus; (4) relativegingival recession (RGR), measured asthe distance from the gingival margin tothe incisal border of the tooth; (5) rela-tive clinical attachment level (CAL) asPD1RGR; (6) non-carious cervicallesion height (CLH), measured as thedistance between the coronal and theapical margins of the non-carious cervi-cal lesion; (7) height of the non-cariouscervical lesion located on the root sur-face (CLH-R): the CEJ was estimated bythe method described by Zucchelliet al. (2006) using digital photographsobtained with a camera positioned per-pendicular to the buccal surface of theexperimental teeth at a magnificationratio of 1:1. The distance from theestimated CEJ to the incisal border ofthe tooth and RGR were measured usingan image analysis software. CLH-R wascalculated by subtracting the distancefrom the estimated CEJ to the incisalborder from RGR. This parameterallowed the calculation of the percen-tage of root coverage. The subtraction ofthe non-carious CLH on the root from

792 Santamaria et al.

r 2009 John Wiley & Sons A/S

the total CLH provided the amount ofcervical lesion located on the crown(CLH-C); (8) keratinized tissue width(KTW), measured as the distance fromthe gingival margin to the mucogingivaljunction; (9) keratinized tissue thick-ness (KTT); (10) DS, which was deter-mined by asking patients about thepresence or absence of cervical sensi-tivity in the sites included in the studybefore and after treatment. No thermalstimulus was applied to assess this para-meter and the patients simply answeredwhether they felt any discomfort in thearea.

The PD was measured using a manualperiodontal probe. The RGR, non-CLH,and KTW were measured using a pair ofdividers and a digital caliper with a0.01mm precision. The KTT was mea-sured using a pierced endodontic sprea-der, perpendicular to a mid-pointlocation between the gingival marginand the mucogingival junction andthrough the soft tissue with light pres-sure until a hard surface was felt. Thesilicone stop was then placed in tightcontact with the external soft tissuesurface. After carefully removing thespreader, penetration depth was mea-sured with a digital caliper. The PD,RGR, CAL, visible plaque at the site(VPS) included in the study and BOPwere measured at baseline and 45 days,and 2, 3 and 6 months after surgery. TheKTW and KTT were obtained at base-line and at 6 months post-operatively.

Before the beginning of the study, theexaminer (M. P. S.) measured the PDand RGR of all patients, two times,within 24 h, with at least 1 h betweenthe examinations. The examiner wasjudged to be reproducible after fulfillingthe pre-determined success criteria. Thek index was calculated to PD, resultingin 91% reproducibility, and the intra-class correlation was calculated relativeto gingival recession, resulting in 89%agreement. The masking of the exam-iner was not practical, because it waspossible to observe whether the glassionomer restoration was applied at thesite. Thus, it was impossible to hide,which treatment each site received.

Surgical procedures

All the surgical procedures were carriedout by one operator (E. A. S.). The siteswere randomly assigned by flipping acoin (F. F. S.) to the control group or thetest group immediately before surgery.The control group received CTG group

and the test group was subjected to CTGplus a resin-modified glass ionomerrestoration (CTG1R group).

Briefly, after local anaesthesia (lido-caine with 1:100,000 epinephrine DFL,Rio de Janeiro, RJ, Brazil), an intra-sulcular incision was made at the buccalaspect of the involved tooth. Two hor-izontal incisions were made at rightangles to the adjacent inter-dental papil-lae, 1mm apical to the level of thecoronal border of the non-carious cervi-cal lesion, without interfering with thegingival margin of neighbouring teeth.Two oblique vertical incisions wereextended beyond the mucogingivaljunction and a trapezoidal mucoperios-teal flap was raised up to the muco-gingival junction. After this point, asplit-thickness flap was extended api-cally, releasing the tension and favour-ing coronal positioning of the flap. In theCTG group, the root and non-cariouscervical lesion were planed with a fin-ishing bur (KG Sorensen, 9803FF, SP,Brazil) and curettes until the tooth sur-face became smooth. In the CTG1Rsites, a sterile rubber dam was placedto isolate the operative field and the non-carious cervical lesion restoration wasperformed with resin-modified glassionomer cement (Vitremer, 3M ESPE,Saint Paul, MN, USA), following themanufacturer’s instructions. The resto-ration was performed in order to rees-tablish the entire defect caused by thecervical wear. The entire length of thenon-carious cervical lesion was restoredand the original contour of the tooth wasrestored. Afterwards, the epithelium onthe adjacent papillae was stripped awayand the CTG harvested from the palateusing a scalpel with parallel blades(1.5mm distant from each other) wasplaced in such a way as to cover theentire non-carious lesion (CTG controlgroup) or the restoration (CTG1R testgroup). Then the flap was coronallypositioned and sutured (6.0 Polygalactin910, Ethicon Inc., Sao Jose dos Campos,Brazil) to completely cover the graft.

Post-operative care

Patients were instructed to take analge-sics (500mg sodium dipyrone every 6 hfor 2 days) and were instructed to dis-continue toothbrushing around the sur-gical sites during the initial 30 days aftersurgery. During this period, plaquecontrol was achieved with a 0.12%chlorhexidine solution rinse used twicea day. After this period, gentle tooth-

brushing with soft-bristle toothbrushwas allowed.

Sutures were removed after 7 daysand the patients were enrolled in aperiodontal maintenance programme(professional plaque control and oralhygiene instruction) weekly during thefirst month and monthly during the 6months.

Statistical analysis

Descriptive statistics were expressed asmean ! standard deviation (SD). ThePD, RGR and relative CAL were exam-ined by the Friedman’s test to evaluatedifferences within groups, followed by apost hoc non-parametric test for multi-ple comparisons and by the Mann–Whitney test to evaluate differencesbetween groups. The CLH (CLH),height of the non-carious cervical lesionlocated on the root (CLH-R) and on thecrown (CLH-C) surfaces were examinedby the Mann–Whitney test to evaluatedifferences between groups. The DS, thevisible plaque at the site included in thestudy (VPS) and the BOP at the siteincluded in the study (BOP) were exam-ined by the w2-square test. The KTWand the KTT were examined by theMann–Whitney test to evaluate differ-ences within and between groups. Asignificance level of 0.05 was adoptedfor all statistical comparisons.

Power calculation

The study power was calculated usingthe SAS 9.01 software (Release 9.1,2003, SAS Institute Inc., Cary, NC,USA). This analysis indicated that with16 subjects in each group, the studywould have 480% power to detect a1mm difference (1.0mm between CTGand CTG1R groups was considered asclinically significant) in RGR and CALbetween the two groups. After the com-pletion of the study, considering the SDof each group of the present study, thepower values were confirmed to be480% to detect a 1mm difference inRGR and CAL between the two groups.A difference of 1.0mm between CTGand CTG1R groups was considered asclinically significant.

Results

Healing was uneventful for all patientsand none were excluded from the study(40 patients one defect in each patient –

Connective tissue graft on restored root surface 793

r 2009 John Wiley & Sons A/S

total of 40 defects treated). Full-mouthPI and FMBOP were maintained below20%, indicating a good standard ofsupragingival plaque control during thestudy period. The sites included in thestudy did not show BOP or visibleplaque during the entire study period.A flow diagram of the participants in thestudy is depicted (Fig. 1). Table 1 showsthe patients’ characteristics at baseline.No adverse event was observed in anypatient during the study.

Cervical lesion

The mean CLH was 3.22 ! 0.52mm forthe CTG group and 3.27 ! 0.68mm forthe CTG1R group (p40.05). Using themethod described by Zucchelli et al.(2006), it was possible to estimate theplace where the lost CEJ was located.Consequently, it was possible to identifythe total amount of root (CLH-R) andcrown (CLH-C) affected by the non-carious cervical lesion. CLH-R was2.45 ! 0.53 for the CTG group and2.36 ! 0.71 for the CTG1R group,representing 74.88 ! 8.66% and 70.76! 9.81% of the total CLH, respectively.CLH-C was 0.77 ! 0.26mm for theCTG group and 0.91 ! 0.23mm forthe CTG1R group. The differencesobserved between groups were not sta-tistically significant for these parameters(p40.05).

Gingival recession

The two groups presented statisticallysignificant reductions in the RGR; 2.53! 0.78mm for the CTG group and2.31 ! 0.74mm for the CTG1R group(po0.05). These reductions in the RGRrepresent 77.59 ! 20.15% of the CLHcovered by CTG and 70.0 ! 13.85% byCTG1R. This difference betweengroups was not statistically significant(p40.05) for this parameter. Three sitesin the CTG group and four sites in theCTG1R group had achieved completeCLH coverage after 6 months of obser-vation.

The percentage of root coverage atthe end of the study period was calcu-lated. The CTG group showed a meanroot coverage of 91.91 ! 17.76% andthe CTG1R group showed a mean rootcoverage of 88.64 ! 11.9%. The differ-ence between groups was not statisti-cally significant (p40.05). Table 2shows the characteristics of the cervicallesion in each group and the totalamount of coverage achieved and Figs.

2–10 show the pre-operative view andthe 6-month post-operative outcome.

PD and CAL

The two groups presented statisticallysignificant increases in the PD from thebaseline until the 6-month follow-up.In the CTG group, this parameterchanged from 1.15 ! 0.48 to 2.1 !0.55mm (po0.05), and from 1.1 !0.44 to 2.15 ! 0.67mm (po0.05) inthe CTG1R group. The differencebetween groups was not statisticallysignificant (p40.05).

After 6 months, both groups showedstatistically significant changes frombaseline for CAL: 1.58 ! 0.74mm for

the CTG group (po0.05) and 1.26 !0.9mm for the CTG1R group (po0.05). The difference between groupswas not statistically significant (p40.05)(Table 3).

Keratinized tissue

The two groups produced statisticallysignificant changes in the (KTT and theKTW. The keratinized thickness gainwas 1.03 ! 0.43mm for the CTG groupand 1.1 ! 0.32mm for the CTG1Rgroup, while the keratinized width gainwas 0.67 ! 33 and 0.8 ! 0.4mm,respectively. There was no statisticallysignificant difference between groupseither in KTT or in KTW. Table 4 shows

Fig. 1. Flowchart for the study patients. CTG, connective tissue graft group; CTG1R,connective tissue graft plus restoration group.

Table 1. Patient characteristics at the baseline (N5 40)

CTG CTG1R

Age 23–55 (mean age 31.8 ! 12.2 years) 19–71 (mean age 39.4 ! 20.4 years)Gender 10 males and 10 females 11 males and nine females

Canines: 8 (40%) Canines: 10 (50%)Teeth 1PM: 9 (45%) 1PM: 6 (30%)

2PM: 3 (15%) 2PM: 4 (20%)FMPI 19.4% 18.5FMBI 18% 14%

FMPI, full-mouth visible plaque index; FMBI, full-mouth sulcus bleeding index; CTG, connective

tissue graft; CTG1R, connective tissue graft plus restoration.

Table 2. Mean values and standard deviation for CTG and CTG1R

CTG CTG1R p-value

CLH 3.22 ! 0.52mm 3.27 ! 0.68mm 0.81CLH-R 2.45 ! 0.53mm

(74.88 ! 8.66%)2.36 ! 0.71mm

(70.76 ! 9.81%)0.43

CLH-C 0.77 ! 0.26mm 0.91 ! 0.23mm 0.67CLH coverage 77.59 ! 20.15% 70.0 ! 13.85% 0.2Root coverage 91.91 ! 17.76% 88.64 ! 11.9% 0.74

CLH, non-carious cervical lesion height; CLH-R, non-carious cervical lesion height located on the

root; CLH-C, non-carious cervical lesion height located on the crown; CTG, connective tissue graft;CTG1R, connective tissue graft plus restoration.

p-values were calculated by the Mann–Whitney test to evaluate differences between groups.

794 Santamaria et al.

r 2009 John Wiley & Sons A/S

the mean and SD of PD, CAL, RGR,KTT and KTW of the test and controlgroups.

Bleeding on probing

FMBI remained low during the entirestudy period. FMBI was 18% for CTGand 14% for CTG1R at baseline and16.7% and 16% at the 6-month evalua-tion, respectively (p40.05). Addition-ally, low levels of FMPI were observedduring the entire study period: mean of

19.4% for CTG and 18.5% for CTG1R.No BOP was observed at any siteincluded in the study in any evaluatedperiod.

Fig. 3. Probing showing the depth of thenon-carious cervical lesion.

Fig. 2. Pre-operative view of the connectivetissue graft plus restoration group site.

Fig. 4. Isolation of the operative field afterthe flap was raised. Note that the entirelength of the non-carious cervical lesionwas included.

Fig. 5. Lateral view of the same tooth ofFig. 4, now restored. Note that the entirenon-carious cervical lesion was restored.

Fig. 6. Connective tissue graft positioned.The connective tissue graft was positionedin order to cover the entire restoration in theconnective tissue graft plus restoration groupand the entire non-restored cervical lesion inthe connective tissue graft group.

Fig. 7. Final suture showing that the con-nective tissue graft was completely covered.

Fig. 8. Clinical outcome after 6 months ofthe connective tissue graft plus restorationgroup site.

Fig. 9. Pre-operative view of the connectivetissue graft site (CTG group).

Connective tissue graft on restored root surface 795

r 2009 John Wiley & Sons A/S

DS

In the study sample, 60% of the subjects(12 patients) from the CTG group and70% of the subjects (14 patients) fromthe CTG1R group reported DS atbaseline. After 6 months, the CTGgroup presented 35% (seven patients)of sites exhibiting this symptom andthe CTG1R group showed 5% (onepatient). The reduction in the percentageof sites with DS was statistically sig-nificant for both groups (po0.05), and astatistically significant difference forthis parameter was observed betweengroups (po0.05). Figure 11 shows thereduction in DS.

Discussion

Because gingival recession is frequentlyassociated with cervical wear, someprevious clinical trials have evaluatedthe ability of the coronally advancedflap to cover this combined lesion (San-tamaria et al. 2008). The comparisonbetween restored and non-restored siteshas been performed previously (Santa-maria et al. 2008, Lucchesi et al. 2007).However, there is a need for clinicaltrials evaluating different approaches todeal with this common condition. In thepresent study, the CTG was used to treatgingival recession, associated with anon-carious cervical lesion alone orcombined with a glass ionomer restora-tion of the cervical wear. Therefore, thepresent study evaluated the ability of theCTG to cover the combined defect(CTG group) and determined whetherthe glass ionomer restoration interfereswith the amount of coverage achievedby the CTG (CTG1R group).

The observed changes in the RGRafter 6 months were 2.53 ! 0.78 and2.31 ! 0.74mm for CTG and CTG1R,

respectively (p40.05). These changesin the position of the gingival margin toa more coronal level provided a compar-able percentage of CLH coverage(77.59 ! 20.15% in the CTG groupand 70.0 ! 13.85% in the CTG1Rgroup, p40.05) and gain of CAL(1.57 ! 0.74% in the CTG group and1.26 ! 0.9% in the CTG1R group,p40.05) after the two treatmentapproaches. Therefore, it can beassumed that the presence of the restora-tion on the cervical area may not preventthe amount of soft tissue coverage thatcan be achieved by a CTG flap in thissituation, considering the period ofobservation of 6 months.

The values of CLH coverage reportedin the present study are related to the

total height of the cervical lesion (crownand root zones). Therefore, the CLHcoverage reported in the present studyshould not be directly compared withother studies that included gingivalrecession on intact roots. This compar-ison is not possible because the non-carious cervical lesion simultaneouslyaffects parts of the root and crown ofthe tooth and with its progression, theCEJ generally disappears. A new line isestablished, coronal to the original CEJ,representing the incisal border of thenon-carious cervical lesion, and is oftenmistaken for the CEJ (Zucchelli et al.2006). Only the part of the non-cariouscervical lesion, located on the root couldbe predictably covered by soft tissueafter the surgical procedure. This is

Table 3. Mean gain in CAL and RGR at 6 months (mm)

CTG CTG1R p-value

CAL gain 1.58 ! 0.74 1.26 ! 0.9 0.16RGR reduction 2.53 ! 0.78 2.31 ! 0.74 0.41

CTG, connective tissue graft group; CTG1R, connective tissue graft plus resin-modified glass

ionomer restoration group; CAL, clinical attachment level; RGR, relative gingival recession.p-value were calculated by the Mann–Whitney test to evaluate differences within groups.

Table 4. Clinical results in mm (mean ! SD; n5 40 patients)

Baseline 45 days 2 months 3 months 6 months

PD CTG1R 1.1 ! 0.44 1.9 ! 0.64n 2 ! 0.56n 2 ! 0.56n 2.15 ! 0.67n

CTG 1.15 ! 0.48 1.98 ! 0.6n 2 ! 0.45n 2.15 ! 0.48n 2.1 ! 0.55n

NIC CTG1R 12.89 ! 1.09 11.4 ! 1.28n 11.51 ! 1.15n 11.57 ! 1.12n 11.63 ! 1.08n

CTG 12.85 ! 2.06 11.1 ! 1.84n 11.15 ! 1.72n 11.27 ! 1.7n 11.27 ! 1.17n

RGR CTG1R 11.79 ! 1.09 9.5 ! 0.87n 9.51 ! 0.88n 9.57 ! 0.81n 9.48 ! 0.82n

CTG 11.7 ! 2.01 9.12 ! 1.55n 9.15 ! 1.46n 9.12 ! 1.52n 9.17 ! 1.53n

KTT CTG1R 0.85 ! 0.19 – – – 1.95 ! 0.42n

CTG 0.9 ! 0.23 – – – 1.93 ! 0.53n

KTW CTG1R 2.54 ! 1.17 – – – 3.34 ! 0.91n

CTG 2.38 ! 1.22 – – – 3.05 ! 1.11n

nStatistically significant difference within groups (po0.05) by the Friedman’s test.wStatistically significant difference between groups (po0.05) by the Mann–Whitney test.CTG1R, connective tissue graft plus resin-modified glass ionomer restoration group; CTG,

connective tissue graft group; PD, probing depth; RGR, relative gingival recession; KTT, keratinized

tissue thickness; KTW, keratinized tissue width.

Fig. 11. Percentage of dentin sensitivity (DS) occurrence. The two groups presented astatistically significant reduction in DS after the treatment.

Fig. 10. Clinical outcome after 6 months ofthe connective tissue graft site.

796 Santamaria et al.

r 2009 John Wiley & Sons A/S

probably the reason why no total CLHcoverage could be observed previouslywith the coronally advanced flap (San-tamaria et al. 2008). However, in thepresent study, a total of seven sites(three in CTG group and four in theCTG1R group) presented completeCLH coverage. One possible explana-tion for this result could be that thepresence of the CTG beneath the flapmight have prevented the collapse of theflap inside the dead space created by thecervical lesion. The presence of theconnective tissue under the flap mighthave provided an adequate support tothe flap and, as a consequence, providedbetter stability (Mele et al. 2008). Suc-cessful cases in which the gingival mar-gin was moved coronally, beyond theCEJ, using the CTG were shown byMcNeelly (2005). This is probably thereason why both groups of the presentstudy showed slightly better averagesof CLH coverage, when compared withthe groups of the previous study (Santa-maria et al. 2008). However, additionalstudies are necessary to test this hypo-thesis.

In order to explore the hypothesis thatthe most coronal zone of the non-cariouscervical lesion was mainly composed bythe crown portion of the lesion, anestimation of the position of the CEJby the method described by Zucchelliet al. (2006) was performed. Accordingto this method, a scalloped line thatrepresents the CEJ lost is constructedfollowing the patient biotype and con-necting the ideal dimension of the adja-cent papilla. Therefore, it was possibleto estimate the part of the CLH locatedon the root (CLH-R). The CLH-R was2.45 ! 0.53mm (74.88 ! 8.66% of theCLH) for the CTG group and 2.36 !0.71mm (70.76 ! 9.81% of the CLH)for the CTG1R group. Based on thesevalues, mean root coverage (CLH-Rcoverage) was calculated, reaching91.91 ! 17.76% for the CTG groupand 88.64 ! 11.9% for the CTG1Rgroup (p40.05). The mean values ofroot coverage observed in the presentstudy are comparable to the onesreported in other studies for this proce-dure (Allen & Miller 1989, Harris &Harris 1994, Wennstrom & Zucchelli1996, Pini-Prato et al. 2000, Cortelliniet al. 2009). However, caution should beexercised due to the subjective compo-nent of the method used to estimate theCEJ in the present study, which differsfrom the direct measurement obtained instudies with intact roots.

An interesting finding of the presentstudy is related to BOP. In spite of thesubgingival location of the apical mar-gin of the restoration, as a consequenceof the coverage achieved after thesurgical procedures, no site in eitherthe CTG1R group or in the CTG groupshowed BOP. This result is in accor-dance with other studies (Lucchesi et al.2007, Santamaria et al. 2007, 2008,2009). Conversely, these data do notagree with studies that demonst-rated that the presence of restorationmargins close to the gingival marginor within the crevicular space may causegingival inflammation (Larato 1972).Dragoo (1996), Dragoo (1997) andAlkan et al. (2006) studies showedthat periodontal health was maintainedwhen a resin-modified glass ionomerwas used for subgingival or transgingi-val restorations. Therefore, the selectionof the resin-modified glass ionomerto be used in the present study wasbased on the results of these previousstudies. The suggested biocompati-bility of the material, added to the factthat the patients were followed upmonthly for prophylaxis, plaque controland oral hygiene instructions, mayhelp to explain the gingival healthobserved during the study. In addition,flap elevation allowed proper isola-tion of the operative field and a well-finished filling could be achieved,which might have facilitated plaquecontrol.

In the present study, the patients wereasked about the presence of dentinhypersensitivity (DS) before and aftertreatment, without application of anythermal or tactile stimuli to detectthe sensitivity. The success of the ther-apy should be based on patients’ eva-luation of this symptom (Roccuzzoet al. 2002, Cairo et al. 2008). Theresults revealed a statistically signifi-cant reduction in DS between baselineand the subsequent observation periodsfor the CTG1R group, whereas thesewithin-group differences were notobserved for the CTG group. The com-parison between the groups revealeda statistically significant difference,with better outcomes for the CTG1Rgroup in all post-operative periods. Thismay be related to the fact that mostof the cervical lesions did not achievecomplete coverage with gingival tissue.Thus, part of the cervical lesion wasstill exposed to the oral environmentin the CTG group. Conversely, cervicallesions in the CTG1R group were

restored, sealing the exposed dentinaltubules and reducing the chancesof symptoms. Again, the subjec-tive nature of DS evaluation in thepresent study should be pointed out.A decision was made during the plan-ning of the study to limit this evaluationto a simple question, without the useof a scale. If the patient reported anysensitivity, regardless of the intensity,it was considered positive for the ana-lysis.

The PD showed a statistically signifi-cant change between baseline and post-operative periods in the two groups.They showed an increase of about1mm in the PD after the treatments.Although the increase was statisticallysignificant, the clinical importance ofthis alteration could be questionedbecause both groups presented shallowPD after 6 months: 2.1 ! 0.55mm forthe CTG group and 2.15 ! 0.67mmfor the CTG1R group. The differ-ences between groups were not sta-tistically significant at any period ofrevaluation.

Within the limits of this short-termstudy, it can be concluded that thepresence of resin-modified glass iono-mer restoration may not interfere withthe percentage of soft tissue coverage,when a CTG is performed for the treat-ment of Miller Class I gingival reces-sions, associated with non-cariouscervical lesions. The combined treat-ment showed better results in the reduc-tion of DS. However, these conclusionsshould be interpreted with caution,based on the following considerations:the periodontal surgery associated withthe restorative procedure required alonger clinical time, compared with theisolated surgical procedure. Addition-ally, no assessment of patient satisfac-tion using a standardized approach wasperformed in the present study. Thestatistical analysis of the present studyincluded a power value 480% to detecta clinically significant difference of1.0mm between CTG and CTG1R inthe RGR and CAL. Although this is anacceptable value, further studies withlarger samples are strongly recom-mended to confirm these results. Long-itudinal observation is also necessary toevaluate the stability of the results andto establish the long-term success ofthis combined approach. Other restora-tive materials and surgical techniquesshould be tested to achieve the bestcombination to treat this particular com-bined lesion.

Connective tissue graft on restored root surface 797

r 2009 John Wiley & Sons A/S

Acknowledgement

The authors would like to thank theFoundation for the Development of Per-sonnel in Higher Education – CAPESfor supporting Dr. Mauro P. Santamaria.Dr. Enilson A. Sallum wants to thank thesupport of CNPq – ‘‘Conselho Nacionalde Desenvolvimento Cientıfico e Tecno-logico’’ (CNPq #300817/2007-0).

References

Ainamo, J. & Bay, I. (1975) Problems andproposals for recording gingivitis and plaque.International Journal of Periodontics andRestorative Dentistry 25, 229–235.

Alkan, A., Keskiner, I. & Yuzbasioglu, E.(2006) Connective tissue grafting on resinionomer in localized gingival recession. Jour-nal of Periodontology 77, 1446–1451.

Allen, E. P. & Miller, P. D. Jr. (1989) Coronalpositioning of existing gingiva: short termresults in the treatment of shallow marginaltissue recession. Journal of Periodontology60, 316–319.

Bartlett, D. W. & Shah, P. (2006) A critical reviewof non-carious cervical (wear) lesions and therole of abfraction, erosion, and abrasion. Jour-nal of Dental Research 85, 306–312.

Cairo, F., Pagliaro, U. & Nieri, M. (2008)Treatment of gingival recession with coron-ally advanced flap procedures: a systematicreview. Journal of Clinical Periodontology35 (Suppl. 8), 136–162.

Cortellini, P., Tonetti, M., Baldi, C., Francetti,L., Rasperini, G., Rotundo, R., Nieri, M.,Franceschi, D., Librriola, A. & Pini Prato,G (2009) Does placement of a connectivetissue graft improve the outcomes of coron-ally advanced flap for coverage of singlegingival recessions in upper anterior teeth?A multi-centre, randomized, double-blind,clinical trial. Journal of Clinical Perio-dontology 36, 68–79.

Dragoo, M. R. (1996) Resin-ionomer andhybrid-ionomer cements: part I. Comparisonof three materials for the treatment of sub-gingival root lesions. International JournalPeriodontics and Restorative Dentistry 16,594–601.

Dragoo, M. R. (1997) Resin-ionomer andhybrid-ionomer cements: part II, human clin-ical and histologic wound healing responsesin specific periodontal lesions. InternationalJournal of Periodontics and Restorative Den-tistry 17, 75–87.

Harris, R. J. & Harris, A. W. (1994) Thecoronally positioned pedicle graft with inlaidmargins: a predictable method of obtainingroot coverage of shallow defects. Interna-tional Journal of Periodontics and Restora-tive Dentistry 14, 229–241.

Larato, D. C. (1972) Influence of a compositeresin restoration on the gingival. ProstheticDentistry 28, 402–404.

Litonjua, L. A., Andreana, S., Bush, P. J.,Tobias, T. S. & Cohen R, E. (2003) Non-carious cervical lesions and abfractions: are-evaluation. Journal of American DentalAssociation 134, 845–850.

Lucchesi, J. A., Santos, V. R., Amaral, C. M.,Peruzzo, D. C. & Duarte, P. M. (2007) Cor-onally positioned flap for treatment of restoredroot surfaces: a 6-month clinical evaluation.Journal of Periodontology 78, 615–623.

McNeely, T. E. (2005) Coronally repositioningthe cemento-enamel junction to address gin-gival margin discrepancies. Journal of Perio-dontology 76, 138–142.

Mele, M., Zucchelli, G., Montevecchi, M. &Checchi, L. (2008) Bilaminar technique inthe treatment of a deep cervical abrasiondefect. International Journal of Periodonticsand Restorative Dentistry 28, 63–71.

Muhlemann, H. R. & Son, S. (1971) Gingivalsulcus bleeding – a leading symptom ininicial gingivitis. Helvetica OdontologicaActa 15, 107–113.

Pini-Prato, G., Pagliaro, U., Baldi, C., Nieri, M.,Saletta, D., Cairo, F. & Cortellini, P. (2000)Coronally advanced flap procedure for rootcoverage. Flap with tension versus flap with-out tension: a randomized controlled clinicalstudy. Journal of Periodontology 71, 188–201.

Roccuzzo, M., Bunino, M., Needleman, I. &Sanz, M. (2002) Periodontal plastic surgeryfor treatment of localized gingival recessions:a systematic review. Journal of ClinicalPeriodontology 29, 178–194.

Sangnes, G. & Gjermo, P. (1976) Prevalence oforal soft and hard tissue lesions related tomechanical toothcleansing procedures. Com-munity Dentistry and Oral Epidemiology 4,77–83.

Santamaria, M. P., Feitosa, D. S., Nociti , F. H.Jr., Casati, M. Z., Sallum, A. W. & Sallum, E.A. (2009) Cervical restoration and theamount of soft tissue coverage achieved bycoronally advanced flap. A 2-year follow-uprandomized controlled clinical trial. Journalof Clinical Periodontology 36, 434–441.

Santamaria, M. P., Suaid, F. F., Nociti, F. H. Jr.,Casati, M. Z., Sallum, A. W. & Sallum, E. A.(2007) Periodontal surgery and glass ionomer

restoration in the treatment of gingival reces-sion associated with a non-carious cervicallesion. Report of three cases. Journal ofPeriodontology 78, 1146–1153.

Santamaria, M. P., Suaid, F. F., Nociti, F. H. Jr.,Casati, M. Z., Sallum, A. W. & Sallum, E. A.(2008) Coronally positioned flap plus resin-modified glass ionomer restoration for thetreatment of gingival recession associatedwith non-carious cervical lesion. A rando-mized controlled clinical trial. Journal ofPeriodontology 79, 621–628.

Wennstrom, J. L. & Zucchelli, G. (1996)Increased gingival dimensions. A significantfactor for successful outcome of root cover-age procedure? A 2-year prospective clinicalstudy. Journal of Periodontology 23, 770–777.

Zucchelli, G., Testori, T. & De Sanctis, M.(2006) Clinical and anatomical factors limit-ing treatment outcomes of gingival recession:a new method to predetermine the line of rootcoverage. Journal of Periodontology 77,714–721.

Supporting Information

Additional Supporting Information maybe found in the online version of thisarticle:

Table S1. Supporting information inaccordance with the CONSORT State-ment 2001 checklist used in reportingrandomized trials.

Please note: Wiley-Blackwell is notresponsible for the content or function-ality of any supporting materials sup-plied by the authors. Any queries (otherthan missing material) should be direc-ted to the corresponding author for thearticle.

Address:Enilson Antonio SallumDepartment of Prosthodontics and PeriodonticsDivision of PeriodonticsPiracicaba Dental SchoolUniversity of Campinas – UNICAMPAv. Limeira901. AreiaoPiracicabaSPBrazilE-mail: easallum@fop.unicamp.br

Clinical Relevance

Scientific rationale for the study:Gingival recession is frequently asso-ciated with a non-carious cervicallesion. The literature lacks controlledstudies evaluating the use of CTGsand restorations to treat this condition.

Principal findings: The present studyshows that CTG alone, or in combi-nation with glass ionomer restora-tion, may provide comparable softtissue coverage in the treatment of agingival recession-associated cervi-cal lesion.

Practical implications: The presentresults suggest that the combinedapproach may be considered as atreatment option for the type of lesionincluded in the study. Long-termobservations are necessary to confirmthe stability of the achieved results.

798 Santamaria et al.

r 2009 John Wiley & Sons A/S