J Periodontol • April 2004

Guided Bone Regeneration May BeNegatively Influenced by NicotineAdministration: A Histologic Study in DogsJuliana B. Saldanha,* Suzana P. Pimentel,* Marcio Z. Casati,* Enilson A. Sallum,* Deise Barbieri,†Heitor Moreno Jr.,† and Francisco H. Nociti Jr.*

565

Background: A series of animal and in vitro data confirms thatnicotine impairs bone healing, diminishes osteoblast function,and causes autogenous bone graft morbidity. Therefore, this studyaimed to investigate the impact of nicotine on the healing of bonedefects treated by the guided bone regeneration (GBR) principle.

Methods: Sixteen mongrel dogs were used. One defect wassurgically created bilaterally and randomly assigned as anexpanded polytetrafluoroethylene (ePTFE) membrane site or anon-membrane control site. The animals were randomly assignedto one of the following groups: group 1, placebo (n = 8) and group2, subcutaneous administration of nicotine (2 mg/kg) twice daily(n = 8). After 4 months, the animals were sacrificed and the spec-imens routinely processed for semi-serial decalcified sections. Theevaluated parameters were bone height, bone width, bone density,and bone area of newly formed bone.

Results: Intergroup analysis (Kruskal-Wallis) showed thatmembrane-protected defects in the placebo group demonstratedan increased bone area when compared to membrane-protecteddefects in the nicotine group and non-membrane sites, regardlessof nicotine administration (P <0.05). In addition, nicotine adminis-tration significantly affected bone density in membrane- andnon–membrane-protected sites (P <0.05).

Conclusions: Within the limits of the present study, nicotinemight affect, but not prevent, bone healing in defects treated byguided bone regeneration. The mechanisms of this effect shouldbe investigated further. J Periodontol 2004;75:565-571.

KEY WORDSAnimal studies; guided bone regeneration; nicotine/adverseeffects; smoking/adverse effects; tobacco/adverse effects;wound healing.

* Department of Prosthodontics and Periodontics, Division of Periodontics, Schoolof Dentistry at Piracicaba, University of Campinas, Piracicaba, São Paulo, Brazil.

† Department of Pharmacology, Faculty of Medical Sciences, University of Campinas,Campinas, São Paulo, Brazil.

Treatment of alveolar ridge atrophyand large, isolated ridge defectsconstitutes a significant challenge in

esthetics and implant placement. Numer-ous techniques have been designed toaddress these situations including auto-genous bone grafting, use of bone substi-tutes, and, most recently, guided boneregeneration (GBR). The predictability oftreatment varies with the technique em-ployed, with success dependent upon thehealing potential of the individual.

Smoking has an adverse effect on bonein general. Post-menopausal women whosmoke lose significantly more corticalbone and have more spinal osteoporosisthan their non-smoking counterparts.1

Cigarette smoking may present a delete-rious effect on healing of open tibia-shaftfractures,2 and interfere with osteoblasticfunction.3,4 A 15-patient clinical studyrevealed that 80% of the individuals withimpaired osseous healing were smokers.5

These physiologic responses to smok-ing may be caused by nicotine, which isthe major constituent of the particulatephase of tobacco smoke and its mostcytotoxic and vasoactive substance. In vivoand in vitro studies have demonstratedthat nicotine may inhibit revasculariza-tion of bone grafts,3 negatively impactbone healing,6 and inhibit expression of awide range of cytokines including thoseassociated with neovascularization andosteoblast differentiation.7

Currently, there is limited informationregarding the effect of cigarette com-pounds, such as nicotine, on the healing

30103.qxd 4/12/04 2:57 PM Page 565

566

Guided Bone Regeneration and Nicotine Volume 75 • Number 4

‡ Pentabiótico, Wyeth-Whitehall Ltda, São Paulo, SP, Brazil.§ Gore-Tex Augmentation Material [GTAM], W.L. Gore & Associates, Inc.,

Flagstaff, AZ.� Sigma Chemical Co., St. Louis, MO.

Figure 1.Clinical appearance of a surgically created bone defect before (A) and after (B) membrane placement. Note that stainless-steelscrews stabilized the membranes.

process of bone defects treated by the GBR principle.Therefore, the present study aimed to investigate theimpact of nicotine on bone healing following guided boneregeneration in dogs.

MATERIALS AND METHODSAnimalsSixteen healthy mongrel dogs (approximately 15 kgbody weight; 2 years old) were used in this study. Theanimals were kept in individual cages with access tofood and water ad libitum. Prior to the surgical proce-dures, all animals were allowed to acclimate to the facil-ity environment for 7 days. The protocol was approvedby the University of Campinas Institutional Animal Careand Use Committee.

Surgical ProceduresOral prophylaxis was performed 2 weeks prior to toothextraction and again 2 weeks prior to creation of thebone defects. The animals received 1.5 ml/10 kg ofacepromazine followed by intravenous injection of 25%sodium thiopental solution (0.5 ml/kg) and local admin-istration of 2% xylocaine (1:50,000 epinephrine) in allsurgical procedures. At the beginning of the experiment,the first, second, and third mandibular molars wereremoved bilaterally, creating an edentulous area in theposterior region of the lower jaw. Prophylactic antibiotictreatment‡ was initiated the day prior to defect creationand treatment, and continued for 7 days postoperatively.After 3 months of healing, full-thickness flaps were ele-vated, and one rectangular bone defect was surgicallycreated bilaterally in the area of the previously extractedteeth using a low-speed rotary and hand instruments.Copious sterile saline irrigation was used during prepara-tion of the defects. Although individual differences inalveolar ridge dimensions did not allow a perfect stan-dardization of the defects, a careful effort was made tokeep the defect dimensions reasonably consistent.

Defect dimensions were approximately 8 mm apico-coronal, 12 mm mesio-distal, and 8 mm bucco-lingualat the bottom of the defects (Fig. 1A). Each animalrandomly received a standard expanded polytetrafluo-roethylene (ePTFE) membrane,§ or no membrane (con-trol site). The membranes were trimmed and drapedover the ridge so that they completely covered thedefects and extended beyond the margins of the defectby at least 2 to 3 mm. The membranes were stabilizedand secured in place by stainless-steel fixation screws.After securing the lingual side and prior to securing thebuccal side of the membrane, bleeding in the defectarea was promoted by scraping the bone surface witha sharp cutting instrument. The buccal side of the mem-brane was then positioned and secured in place by twoadditional screws (Fig. 1B). Subsequently, primarywound closure was achieved with vertical mattress andinterrupted ePTFE sutures.§ Oral prophylaxis was per-

formed every 2 weeks. Chlorhexidine rinse was useddaily throughout the experimental phase, and the dogswere maintained on a soft diet. Sutures were removed10 days after surgery.

Subcutaneous nicotine administration� was started theday after defect creation and treatment and continuedthroughout the experimental period in group 2 animals(n = 8). A total dose of 4 mg/kg/day was used, admin-istered twice a day in 12-hour intervals (2 mg/kg eachtime). Group 1 animals (n = 8) received a placebo.

Nicotine and Cotinine Serum Levels: AnalyticalMethodsBlood samples were taken hourly from 15 minutes to 8hours after the first injection of the day on the first andlast days of nicotine administration. Serum samples wereassayed for concentrations of nicotine and cotinine byhigh-pressure liquid chromatography (HPLC), composed

30103.qxd 4/12/04 2:57 PM Page 566

567

J Periodontol • April 2004 Saldanha, Pimentel, Casati, et al.

Figure 2.Line chart illustrating nicotine serum levels from 15 to 480 minutesafter the first injection of the day, on the first (A) and last (B) days ofnicotine administration.

of two pumps¶ programmed by a system controller,¶ aUV-Vis detector¶ set at 260 nm, and a reversed-phasecolumn# (150 mm × 4.6 mm internal diameter × 5 µm).The mobile phase consisted of 20 mM dibasic potas-sium phosphate, 20 mM monobasic potassium phos-phate containing 0.1% triethylamine. The pH of thesolution was adjusted to 6.3 with phosphoric acid, andacetonitrile (10%) was added to the final solution. Theflow rate was 1.0 ml/minute. 2-phenylimidazole wasused as an internal standard. All of the reagents used toperform the method were high performance liquid chro-matography grade. Extraction of the samples followedthe methodology previously described by Nakajimaet al.;8 however, the samples were dried under nitrogenat room temperature. The injection volume was 20 µl,and the limit of quantification was 10 ng/ml.

Histometric ProcedureThe animals were sacrificed 4 months after defect crea-tion by induction of deep anesthesia, with subsequentintravenous sodium pentobarbital overdose. The jawswere removed and fixed in 4% neutral formalin for48 hours. The specimens were demineralized in a solu-tion of equal parts of 50% formic acid and 20% sodiumcitrate for 90 days. Paraffin semi-serial sections (6 µm)were obtained in a bucco-lingual direction and stainedwith hematoxylin and eosin and Masson’s trichrome.

Using an image analysis system,** the followingmeasurements were obtained for the newly formed bone:bone height (BH), bone width (BW), bone density (BD),and bone area (BA). Bone width was evaluated in threedifferent portions of the defect: 1) apical (closest to thebase of the defect), 2) intermediate, and 3) coronal. Toobtain BD and BA, a grid was applied on the imageand the point counting technique used. Measurementswere averaged to allow intergroup analyses.

Statistical AnalysisThe hypothesis that there was no difference betweenthe groups (with or without nicotine) regarding the eval-uated parameters was tested by an intergroup analysisusing either the one-way analysis of variance (ANOVA)or the non-parametric Kruskal-Wallis test (alpha = 0.05).Pair-wise multiple comparisons were done either byBonferroni t test or Dunn’s test (alpha = 0.05) in casesin which the one-way ANOVA or Kruskal-Wallis testdetected significant differences.

RESULTSSerum Levels of Nicotine and CotinineFollow-up of nicotine and cotinine serum levels overtime demonstrated a similar pattern during the first andlast days of nicotine administration (Figs. 2 and 3). Thehighest serum level of nicotine was noted 15 minutesafter its administration, and a time-dependent decreasewas observed. A significant decrease in the serum level

of nicotine was observed 1 hour after its administration,with a tendency for stabilization after the third hour. Withrespect to serum levels of continine, the highest valuewas observed 1 hour after nicotine administration, witha tendency to stabilize after the fourth hour. Interest-ingly, on the first day of nicotine administration, nicotine/cotinine serum levels were slightly higher than on the lastday of administration. However, for both periods, similarvalues were obtained.

Statistical and Histometric AnalysisStatistical analysis showed that, in the absence of nico-tine, membrane-protected defects presented a signifi-cantly greater increase in bone height (BH) thannon-membrane control groups either with or withoutnicotine administration (P <0.05). In addition, use of themembrane resulted in a significant increase in the newlyformed bone area when compared to both non-membranegroups (i.e., with or without nicotine) (P <0.05). On theother hand, regardless of the presence of membraneor nicotine, no significant differences were observed

¶ LC-10ADvp, Shimadzu Corporation, Tokyo, Japan.# Column Luna, Phenomenex, Torrance, CA.** Image-Pro, Media Cybernetics, Silver Spring, MD.

30103.qxd 4/12/04 2:57 PM Page 567

568

Guided Bone Regeneration and Nicotine Volume 75 • Number 4

Figure 3.Line chart illustrating cotinine serum levels from 15 to 480 minutesafter the first injection of the day, on the first (A) and last (B) daysof nicotine administration.

Figure 4.Bar graph showing average and standard deviation for bone height (BH)and bone width at the apical (BWa), intermediate (BWi), and coronal(BWc) portions of the bone defect. No difference was observed betweenBW for any of the evaluated portions. In contrast, note that BHin the MT ctrl (membrane-treated control) sites presented a significantlyhigher value than N-MT ctrl (non–membrane-treated control) and N-MTnico (non–membrane-treated plus nicotine) sites. MT nico refers tomembrane-treated plus nicotine sites.

Figure 5.Bar graph showing average and standard deviation for bone density(BD) and bone area (BA) of the newly formed bone. Note thatthe N-MT ctrl group (non–membrane-treated control) and MT ctrl group(membrane-treated control sites) presented significantly higher valuesfor BD and BA, respectively. N-MT nico refers to non–membrane-treatedplus nicotine sites; MT nico refers to membrane-treated plus nicotinesites.

regarding bone width (BW) in any of the portions eval-uated (i.e., apical, intermediate, and coronal) (P >0.05).

With respect to the influence that nicotine mightexert on the healing process, intergroup analysisdemonstrated that nicotine administration significantlyinfluenced the proportion of mineralized tissue withinthe limits of newly formed bone (BD) in the membrane-protected and non-membrane defects when comparedto control defects without nicotine (P <0.05). Moreover,nicotine administration significantly decreased the areaof newly formed bone in the membrane-protecteddefects when compared to the placebo group (P <0.05).Figures 4 through 6 illustrate the histometric and his-tological data.

DISCUSSIONDuring the past decade, animal and human studies haveshown guided bone regeneration to be a reliable and pre-dictable technique for achieving favorable clinical andhistological outcomes in the treatment of bone defectswhen compared to non-membrane control sites.9-14

These studies demonstrated that membrane placementimproves osseous healing of bone defects, since compet-ing non-osteogenic soft tissue cells are excluded from the

healing area by the presence of a physical barrier. Moststudies have reported a rate of defect filling varying from55% to 100%, depending on the model and type of eval-uation.10-13,15 This observation was confirmed in the pre-sent investigation, which showed that membrane-treatedsites presented a significant increase in newly formedbone area when compared to non-membrane controlsites. In addition, as previously reported,13 the newlyformed bone showed a tendency to present lower density

30103.qxd 4/12/04 2:57 PM Page 568

569

J Periodontol • April 2004 Saldanha, Pimentel, Casati, et al.

Figure 6.Sections illustrating the histological findings for all four groups. A and B illustrate placebo-, and C and D nicotine-treatedgroups without and with membranes, respectively. Note that membrane placement (M) resulted in a higher area of newlyformed bone (NB) when compared to non–membrane-protected defects. In addition, a less dense bone is observed fornicotine-treated groups.The arrows identify the limit between preexisting and new bone.A very dense periosteum (P) wasobserved in some sections from membrane-protected defects regardless of nicotine administration (Masson’s trichrome;original magnification ×4).

in the membrane-protected sites, as compared to thenon-membrane protected sites, in the animals that didnot receive nicotine (74.19% versus 81.51%, respect-ively). Ohnishi et al.14 reported that the chronologicaldifferences observed between sites treated with or without

GBR might result from the influence that the membranemay exert in isolating the periosteum from the bone cav-ity. However, further investigation is required to supportany theory to explain morphological differences in newlyformed bone in sites treated with or without GBR.

30103.qxd 4/12/04 2:57 PM Page 569

570

Guided Bone Regeneration and Nicotine Volume 75 • Number 4

Clinically, GBR has been widely recommended toprovide an appropriate bone volume for titaniumimplant placement and/or for esthetic reasons. There-fore, factors that may affect the predictability of GBRshould be reviewed carefully. Human and animal stud-ies have shown that smoking significantly affects bonehealing5,16 and that nicotine might be involved in thismechanism/effect.3,6,7 To date, limited information isavailable on the effect of cigarette smoke or nicotineon the healing process following GBR. Thus, using apreviously reported model to create the bone defects,10

the present study was designed to investigate histologi-cally the impact of nicotine administration on GBRoutcome. Our findings demonstrated that nicotinemight significantly affect bone healing in GBR-treateddefects, primarily the density and area of newly formedbone.

The main pharmacological action of nicotine is acti-vation of the sympathetic nervous system. Nicotine alsoacts directly on the small vessels, producing vaso-constriction, systemic venoconstriction, and increasingcoronary vascular resistance.17,18 In normal andmembrane-assisted healing situations, the healingprocesses are modulated by the activity of local cellsand signaling molecules.19 When osseous healingexhausts the localized supply of cells and signaling mol-ecules, renewal is contingent upon vascularity and oper-ational activity of endogenous cells activated during theearly phase of bone repair. Therefore, the blood supplyperforms a critical role during the overall process. Theintense vasoconstrictive effect that nicotine exerts on themicrovasculature may inhibit the angioblastic responseduring revascularization in the healing area, and alsolimit the presence of important factors such as cyto-kines. The fact that, in the present study, nicotine pro-moted an inhibition of bone healing following GBR is inagreement with previous observations of an inhibitionof revascularization of bone grafts,3 lowered healing inparietal bone defects,6 and inhibition of genes directlyrelated to neovascularization and osteoblast differenti-ation.7 Moreover, in vitro studies have demonstrated adose-dependent inhibition of proliferation, extracellularmatrix production, and attachment of human gingi-val fibroblasts, in addition to an increased collagenaseactivity,20,21 a dose-dependent inhibition of attachment,increased activity of alkaline phosphatase, and chemo-taxis of periodontal ligament fibroblasts in the presenceof nicotine.22 On the other hand, nicotine has been asso-ciated with increased alkaline phosphatase production,decreased osteoclastic activity, and enhanced humangingival fibroblast attachment.23,24 Therefore, since thisis the first study to investigate the impact of nicotine onbone healing following GBR in the face of such contra-dictory reports, further studies should be considered tocharacterize the role of nicotine in the bone healingprocess.

To our knowledge, there is limited information onnicotine/cotinine serum levels after subcutaneousadministration of nicotine in dogs. Therefore, in thepresent study, care was taken to ensure that nicotine/cotinine serum levels were consistent with thosereported for smokers. Serum levels of cotinine alsowere evaluated because it is one of the main metabo-lites of nicotine and, therefore, has also been used asa marker in studies on the effect of smoking on bio-logical events. Blood samples were collected 15 min-utes and hourly (1 to 8 hours) after the first nicotineinjection of the day, on the first and last days of drugadministration. The first and last days were chosen toevaluate whether two injections of 2 mg/kg of nicotinewould reach blood levels of nicotine in the range ofthose in smokers and whether there would be anycumulative effect during the entire experimental period.Although an initial high nicotine concentration wasobtained (15 minutes to 4 hours), data analysis demon-strated that nicotine serum levels in the present studyremained within the range reported for smokers.Furthermore, despite the long experimental period,nicotine serum levels were observed to be similar onthe first and last days of the experiment; therefore,a cumulative effect of daily injections was not observed.

CONCLUSIONIn conclusion, within the limits of this study, nicotineaffected but did not prevent bone healing in defectstreated by guided bone regeneration. The mechanismsand clinical relevance of our findings remain to be inves-tigated.

ACKNOWLEDGMENTSThis study was supported by the Research Foundationfor the State of São Paulo (grants 01/06481-3 and01/06480-7, FAPESP). Dr. Moreno Jr. was supportedby the National Research Committee (301119/97,CNPq - Brazil).

REFERENCES1. Rapuri PB, Gallacher JC, Balhorn KE, Ryschon KL.

Smoking and bone metabolism in elderly women. Bone2000;27:429-436.

2. Harvey EJ, Agel J, Selznick HS, Chapman JR, HenleyMB. Deleterious effect of smoking on healing of opentibia-shaft fractures. Am J Orthop 2002;31:518-521.

3. Daftari TK, Whitesides TE, Heller JG, et al. Nicotineon the revascularization of bone grafts. Spine 1994;19:904-911.

4. Riebel G, Boden S, Whitesides T, Hutton W. The effect ofnicotine on incorporation of cancellous bone graft in ananimal model. Spine 1995;20:2198-2202.

5. Jones J, Triplett R. The relationship between cigarettesmoking to impaired intraoral wound healing: A review ofevidence and implications for patient care. J Oral Max-illofac Surg 1992;50:237-239.

6. Hollinger JO, Schmitt JM, Hwang K, Soleymani P, BuckD. Impact of nicotine on bone healing. J Biomed Mater

30103.qxd 4/12/04 2:57 PM Page 570

571

J Periodontol • April 2004 Saldanha, Pimentel, Casati, et al.

Res 1999;45:294-301.7. Theiss SM, Boden SD, Hair G, Titus L, Morone MA, Ugbo J.

The effect of nicotine on gene expression during spinefusion. Spine 2000;25:2588-2594.

8. Nakajima M, Yamamoto T, Kuroiwa Y, Yokoi T. Improvedhighly sensitive method for determination of nicotine andcotinine in human plasma by high-performance liquidchromatography. J Chromatogr 2000;742:211-215.

9. Dahlin C, Linde A, Gottlow J, Nyman S. Healing of bonedefects by guided tissue regeneration. Plast ReconstrSurg 1988;81:672-676.

10. Schenk RK, Buser D, Hardwick WR, Dahlin C. Healingpattern of bone regeneration in membrane-protecteddefects: A histologic study in the canine mandible. IntJ Oral Maxillofac Implants 1994;9:13-29.

11. Kostopoulos L, Karring T. Augmentation of the rat mandi-ble using guided tissue regeneration. Clin Oral ImplantsRes 1994;5:75-82.

12. Buser D, Dula K, Hirt HP, Schenk RK. Lateral ridgeaugmentation using autografts and barrier membranes:A clinical study with 40 partially edentulous patients.J Oral Maxillofac Surg 1996;54:420-432.

13. Weng D, Hurzeler MB, Quinones CR, Ohlms A, CaffesseRG. Contribution of the periosteum to bone formation inguided bone regeneration. A study in monkeys. Clin OralImplants Res 2000;11:546-554.

14. Ohnishi H, Fujii N, Futami T, Taguchi N, Kusakari H,Maeda T. A histochemical investigation of the bone for-mation process by guided bone regeneration in rat jaw.Effects of PTFE membrane application periods on newlyformed bone. J Periodontol 2000;71:341-352.

15. Lang NP, Hammerle CH, Bragger U, Lehmann B, NymanSR. Guided tissue regeneration in jawbone defects priorto implant placement. Clin Oral Implants Res 1994;5:92-97.

16. Nociti FH Jr, Cesar-Neto JB, Carvalho MD, Sallum EA,Sallum AW. Intermittent cigarette smoke inhalation mayaffect bone volume around titanium implants in rats.J Periodontol 2002;73:982-987.

17. Isaac PF, Rand MJ. Blood levels of nicotine and physio-logical effects after inhalation of tobacco smoke. Eur JPharmacol 1969;8:269-283.

18. Ray RD. Vascularization of bone grafts and implants.Clin Orthop 1972;87:43-48.

19. Hollinger JO, Wong ME. The integrated processes ofhard tissue regeneration with special emphasis on frac-ture healing. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 1996;82:594-606.

20. Peacock ME, Sutherland DE, Schuster GS, et al. Theeffect of nicotine on reproduction and attachment ofhuman gingival fibroblasts in vitro. J Periodontol 1993;64:658-665.

21. Triton DA, Dabbous MK. Effects of nicotine on proliferationand extracellular matrix production of human gingivalfibroblasts in vitro. J Periodontol 1995;66:1056-1064.

22. Giannopoulou C, Geinoz A, Cimasoni G. Effects of nico-tine on periodontal ligament fibroblasts in vitro. J ClinPeriodontol 1999;26:49-55.

23. Fang MA, Frost PJ, Iida-Klein A, Hahn TJ. Effects of nico-tine on cellular function in UMR 106-01 osteoblast-likecells. Bone 1991;12:283-286.

24. Yuhara S, Kasagi S, Inoue A, et al. Effects of nicotine oncultured cells suggest that it can influence the forma-tion and resorption of bone. Eur J Pharmacol 1999;383:387-393.

Correspondence: Dr. Francisco H. Nociti Jr., Av. Limeira901, Caixa Postal: 052, CEP: 13414-903, Piracicaba, SP,Brazil. Fax: 55-19-3412-5218; e-mail: nociti@fop.unicamp.br.

Accepted for publication August 20, 2003.

30103.qxd 4/12/04 2:57 PM Page 571