J Clin Pcriodanlol 1995: 22: 598-602Printed in Dettmark . AH righls reserved

Copyright © Munk.sgtiard 1995

Clinical periodDntologyISSN 0303-6979

Endodontic pathogens inperiodontal diseaseaugmentation

Leif Jansson^ ^ HelgeLeif Blomlof ' ^ Andrej Weintraub^and Sven Lindskog^^Department of Periodontology, Public DentalService al Skanstull, Stockholm County Cound!^Department of Orai Histoiogy and Celi Bioiogy,Schooi of Dentistry, Karoiinsl<a institutet,Stockholm, Sweden, ^Department of CiinicalBacteriology, Karolinska institutet, i-luddingeUniversity Hospital, Stockholm, Sweden

Jansson L. Ehnevid H, Blomlof L, Weintraub A, Lindskog S: Endodonticpathogens in periodontal disease augmentation. J Clin Periodontol 1995: 22: 598-602. © Munksgaard, 1995.

Abstract, Periapical palhology indicating endodontic infection, when present inperiodontitis-affected teeth, has recently been shown to be correlated to mar-ginal periodontal breakdown. This has been associated with patency of dentinaltubules in the tooth cervix, an area normally devoid of cemenlum followingperiodontal therapy. These studies are, however hampered by that only circum-stantial evidence such as presence of periapical destruction have been appliedas criteria of endodontic infection. The aim of the present investigation was toassess the effects of endodontic pathogens on marginal periodontal wound healingon root surfaces devoid of cementum but surrounded by healthy periodontalmembrane. Significant differences between infected and non-infected teeth werefound with respect to pathological pocket and connective tissue: The experimen-tal defects were covered by approximately 20% more pocket epithelium in infectedteeth while defects in non-infected teeth showed approximately 10% more connec-tive tissue coverage. It was concluded, that an intra-canal infection of endodonticpathogens stimulates epitehal downgrowth along denuded dentin surfaces withmarginal communication. Extrapolated to the clinical situation, endodontic infec-tions in periodontitis-prone patients may augment periodontitis propagation.

Key words: endodcntic pathogens; periodontalhealing; iong epitheiiai junction; monkey;morphometry

Accepted for publication 25 August 1994

The cementum layer covering the per-iphery of the root has been implicatedin preventing spread of intra-canal en-dodontic pathogens to the periodontaltissues (Hamrnarstrom et al. 1986,Ehnevid et al. 1994), Conversely, inareas of dentina! tubules open to theperiodontal space, an endodontic infec-tion may elicit a chronic infiammatoryreaction and maintain progressive ex-ternal root resorption (Hammarstromet al. 1986). Thus, in management oftraumatically injured teeth, the primaryobjective is the prevention of endodon-tic infection in order to preserve toothsubstance which would otherwise belost to external root resorption. How-ever, the role of an endodontic infectionin progression of other periodontalconditions such as marginal peri-odontitis appear less well accepted, al-though recent retrospective clinical

studies have shown a correlation be-tween degree of marginal periodontalbreakdown and pt"esence of periapicalpathology in periodontitis-prone pa-tients (Jansson et al. 1993a, b). The in-fluence of a root canal infection is atleast in the same order of magnitude asoverhanging restorations (Jansson et al.1994). The negative consequences of anendodontic infection on marginal peri-odontal healing following non-surgicaltreatment appear also significant(Ehnevid et al, 1993a, b). Clinical ef-fects of endodontic infections involvedeeper periodonta! pockets, a higherfrequency of vertical destructions andcomparatively more loss of supportingbone.

Extrapolated to an experimental situ-ation, the clinica! effects of an endo-dontic infection may be argued to ini-tially entail the formation of a long epi-

thehal junction (Biomiof et al. 1992)encouraged by margina! p!aque ac-cumulation and infl.ammation (Berg etal. 1990). Although not of greater valueas attachment, the long epitheha! junc-tion may arguabiy prevent widespreadinflammatory root resorption (Karringet ai, 1984). A previous study has indi-cated that an intra-canal infection, al-though only of non-specific indigenousorigin may contribute to the formationof a long epithelial junction in teeth de-void of marginal cementum (Blomlof etal. 1992). However, the specific experi-menta! model used in this study involv-ing extraction and subsequent re-plantation may have influenced the re-sults and favoured epithehal down-growth. Thus, the aim ofthe present in-vestigation was to assess the effects ofendodontic pathogens on marginalperiodonta! wound healing on root sur-

Endodontic pathogens 599

faces devoid of cemenlum but sur-rounded by healthy periodontal mem-brane.

Material and MethodsExperimental animals

The maxillary premolars and the firstmolars in three adult 3-4 year-old mon-keys (Macaca fascicuiaris) were used inthe experiment. The teeth were fullyerupted. The monkeys were fed withtheir normal pelleted food (Astra Ewos,Sodertalje. Sweden) and fresh fruit dur-ing the experimental period. After 5weeks, the animals were killed by anoverdose of Ketalar"^"^ {50 mg/ml keta-min hydrochloride, 10 mg/kg b/w,. ParkDavid Co. Inc., Morris Plains, NJ,USA) and the teeth with adjacent peri-odontal tissues were prepared for histo-logic evaluation.

Experimentai procedure

After anaesthetising the animals withan intramuscular injection of Ketal-ar^*^, anaemia was secured in the surgi-cal areas by local injection of 2%) hdo-cain hydrochloride - 12.5 //g/mi adren-aline (Xylocain-adrenaline, Astra,Sodertalje, Sweden), The premolars andfirst molar of one randomly selectedmaxillary quadrant in each of the 3monkeys were opened. The pulps of thepalatal roots were instrumented withflies (Hedstrom files, nos. 15^0, Mani,Matsunati Seisakusho, Takanezawa-Machi Tochigi-Ken, Japan) and rinsedwith sterile saline. After drying the rootcanals with paper points (DentalPoints, Johnson-Johnson, Sollentuna,Sweden), a bacterial suspension was in-oculated in the root canals. The bac-terial suspension was composed of amixture of 10" colony forming units ofeach of the following 4 bacterial strains,

• Fusobacterium nucleatum (AmericanType Culture Collection, Rockville,MD. USA, ATCC 10953)

• Streptococcus intermedius (AmericanType Culture Collection, Rockville,MD, USA, ATCC 27335)

• Peptostreptococeus micros (CultureCollection University of Goteborg,Sweden, CCUG 17638)

•Porphyromonas gingivalis (clinical iso-late 629/93)The access cavities in the crowns weresealed with zinc oxide-eugenol cement(IRM, L.D, Caulk, Milford, DE, USA).

Subsequent to the root canal treat-ments, a palatal mucoperiostal flap was

raised after an incision along the gingi-val margin in both maxillary quadrantsin each monkey. The bone covering thepalatal roots of the premolars and thefirst molars was thus uncovered. Thebone together with cementum was re-moved with a bur from the cemento-en-amel junction and 2/3 to 3/4 down thelength of the roots. After etching the ex-posed root surfaces for 20 seconds witha 37% aqueous solution of a-phos-phoric acid (Kebo, Stockholm,Sweden), the root surfaces were fioodedwith sterile saline for 1 min and themuco-periostal flaps repositioned totheir presurgical positions and sutured(Catgut 5-0, SSC, Neujausen am Rein-fall, Switzerland).

hHlstoiogic preparation and evaiuation

After sacrifice, the jaw segments con-taining the two premolars and the firstmolar in each maxillary quadrant weredissected out and fixed in 5% neutral-buffered formalin for 48 h. The tissuespecimens were demineralized in 10%formic acid, infiltrated and embeddedin paraffin and sectioned longitudinallyin a bucco-lingual direction through thecentral parts of the teeth. Each sectionwas 5 /im thick. The sections werestained with hematoxylin and eosin andexamined in a light microscope.

The three central-most sections (ap-proximately 200 iim apart) from the de-fects were selected for classification ofthe periapical healing pattern as well asfor histomorphometrical measure-ments. Ali measurements were madewith a computer based image analysisequipment (Hamamatsu Argus-50,Hamamatsu City, Japan) and per-formed parallel to the long axis of theteeth. The following measurements weremade in the defects,

• Defect length: the distance from thecemento-enamel junction to the api-cal limit of the denuded area.

• Flap coverage: the fiap was measuredfrom apical extent of the defect to itscoronal extent.

• Periodontal pocket: pocket epithel-ium comprizing multi-layered gingi-val epithelium with proliferationsinto the underlying connective tissue(Fig, 1),

• Long epithelial junction: uniformlythin layer of epithelium (1 to 5 eel!layers thick) without any prolifer-ations into the underlying connectivetissue (Fig. 2).

• Connective tissue: connective tissue

Fig, J, Pockel epithelium (arrowheads) com-prising multi-lctyered gingival epitheliumwith proliferaiions into the underlying con-nective tissue and the connective tissue alongthe root surface (arrows).

with collagen fibers oriented pre-dominantly parallel to the root sur-face (Figs. 1, 3),

• Reparative cementum: mineralizedtissue deposited on the root surface.

• Root resorption: resorption cavitieson the root surface (Fig. 3),

• Ankylosis: alveolar bone in contactwith the denuded dentin surface.

The length along the root surface foreach of the healing reactions was ex-pressed in percent of the total length ofthe experimental defects in each sectionand means for infected and non-in-fected groups were calculated.

Statistics

The Wilcoxon rank sum test combiningdata from several blocks (Lehmann1975) was used to test significance ofdifferences between the results from theinfected and non-infected control teeth.Differences were considered significantat /?<0.05.

Results

The monkeys appeared to tolerate theexperiments well and no adverse reac-tions were noted during the observation

600 Jansson et al.

i , 2 Long epithelial junction (arrows) con-sisting of a uniformly thin layer of epilheliumwithout any proliferations into the underly-ing connective tissue. The epithelium is arti-factually separated from the root surface.

period and no teeth were lost. Therewere no statistically significant differ-ences between mean values for the totallengths of the experimental defects ininfected (3.3 mm, SD 0,6) and non-in-fected (3.5 mm, SD 0.9) teeth. Simiiarlythere were non-significant differencesbetween flap coverage between infected(110%, SD 11.3) and non-infected(120%, SD 12,6) teeth. The results ofthe morphometrical analyses are sum-marized in Table I. Significant differ-ences between infected and non-in-fected teeth were found with respect toperiodontai pocket and connectivetissue: The experimental defects werecovered by approximately 20% morepocket epithelium in infected teethwhile defects in non-infected teethshowed approximately lO'Ki more con-nective tissue coverage.

All non-infected teeth and 7 out of 9infected teeth displayed normal pert-apical histology with numerous bloodvessels interlaced with periodontaifibres in a non-inflamed periodontaispace. The periapical area in two of theinfected teeth were found to contain mi-nor granulomas consisting oi a chronic-ally inflamed cell-rich fibrous tissue,

Discussioti

Marginal epithehal root coverage inteeth infected with endodontic patho-gens implicated in periapical diseaseetiology (Sundqvist 1976, Kipioti et al.1984, Kerekes & Olsen 1990, Slots &Taubman 1992) were found to be twicethat of non-infected teeth. Approxi-tnately 50%) of the marginal defects ininfected teeth were covered by epithel-ium (pocket and long junctional epi-thelium) already after 5 weeks indi-cating a significant influence of the en-dodontic infection on marginalperiodontai healing. However, beforethe experimental and ciinical impli-cations of this can be discussed some

methodological aspects need to be ad-dressed.

The monkey as ati experimental ani-mal has been widely employed in bothtraumatoiogical (for review, see Andre-asen & Andreasen (1994)) and marginalperiodontai (for review, see Lindskog &Biomlof (i 994)) heahng studies, A goodcorrelation between chnical data andexperimentai results under the simiiarconditions has been reported with re-spect to factors such as development ofankylosis following trauma and endo-dontic treatment (Cvek et al, 1974,Lengheden 1994, Andreasen & Andre-asen 1994). Furthermore, the marginaldehiscence model in monekys is amodel which easily lends itself to quan-titative studies of marginal heahng (Ny-man & Karring 1979, Biomlof et a!,1995a, b). In order to assess the influ-ence on marginal healing of an intra-canal infection, a quantitative evalu-ation basis is imperative. Consequently,the histomorphometrical recordings inthe present study were based on prede-fined periodontai healing reactions, andmeans for the reactions in each treat-ment group formed the basis for stat-istical anaiysis. In addition, meanlengths of the defects and flap coveragefor each group were compared andfound not statistically different. Thus,comparisons between treatment groupsare justifled. However, in the study ofmarginal preiodontal healing, the mar-ginal dehiscence model may oftenunderestimate the influence of modify-ing factors such as infection since theanimals most often are not susceptibleto periodontai disease (Friskopp &Blomlof 1988).

It is weli known that a marginal in-fection promotes epithelialization of aperiodontai wound (Caton et al. 1980,Caton & Nyman 1980, Berg et al. 1990)and that gingival retraction can be in-duced by marginai infection. It shouldbe noted that the experimental animals

Table I. Periodontai healing patterns in the marginal defects after a 5-week; observation period

Fig. 3. Connective tissue w ith collagen fibersoriented predominantly parallel to the rootsurface (arrowheads). Note the active root re-sorptions (arrows).

Periodontai healing pattern

flap coverageperiodontai pocketlong epithelium junctionconnective tissuereparative cementumroot resorptionankyiosis

Infected teeth

110(1!.3)32,3(17,0)21.6(14,1)33.1 (16,1)6,2(7,1)

24,6 (25.3)6,8 (8,2)

Non-infected teeth

120(12,6)12,1 (11,4)15,3(10,1)43,8(14,])12.9(10.2)22,9 (22.6)15.9(14.1)

P

NS*

NS«

NSNSNS

The figures indicate means of relative surface area in percent (standard deviation), showing apredefined reaction along the exposed root surface, NS denotes a non-signiiicant differencebetween infected and non-infected teeth while * indicates a significant difference al /><0.05.

Endodontie pathogens 601

in the present study did not undergoany hygiene treatment during the heal-ing period and, consequently, epithelia!down-growth from plaque-inducedperiodontal inflammation must have in-fluenced the healing resuit to a large ex-tent. Despite a supposed lack of suscep-tibihty to periodonta! disease of experi-mental anima!s such as those used inthe present study and individua! differ-ences in resistance to infection betweenthe experimental anima!s, a significanteffect of the endodontic infection onepithe!ia!ization of the marginal woundwas found.

Earlier studies have reported differ-ent degrees of susceptibihty to recur-rence of periodontal disease between in-dividuals depending on piaque contro!fo!lowing invasive periodonta! therap-ies, ranging from a stable condition(Know!es et a!. !979, Axe!sson &Lindhe !98!, Pi!i!strom et a!. !983) totota! recurrence (Nyman et al. 1911).A!though these resu!ts were obtainedc!inica!!y in humans, it appears indis-putab!e that they a!! had !ong epithe!ia!junctions (Caton & Nyman !980, Ca-ton et a!. !98O, Bowers et al. 1989).However, it should be noted, as dis-cussed earher, that individual differ-ences in resistance to infection andpiaque control determine the rate withwhich recurrences occur (Seymour &Heasman !992). The present resu!ts in-dicate that an intra-cana! infection inaddition to margina! infection may sig-nificant!y contribute to recurrences fo!-!owing periodonta! therapy provideddentina! tubu!es are open to the peri-odonta! space, Remova! of the ce-mentum as part of both surgica! andnon-surgica! periodonta! therapy inevi-tab!y opens dentinal tubu!es and thusa pathways for spread of a root cana!infection to the periodontal space. Thelaclc of difference between non-infectedand infected teeth with respect to rootresorption may appear to contradictprevious studies in which a root canalinfection was shown to promote exter-na! root resorption (Hammarstrom eta!. !986, Ehnevid et a!. !994). However,the rapid epithe!ia!ization of the root/flap interface in the infected teeth maybe argued to have prevented resorbingce!!s to gain access to the root surface(Karring et a!. !984). Furthermore, the!ac!c of an infectious endodontic influ-ence in the non-infected teeth neitherpromoted epitheiia! downgrowth norexterna! root resorption resu!ting in-stead in connective tissue formation

(Berg et a!. !990). Thus, the net resuitof these different compensatory factorsacting in the two groups wil! eliminatedifferences in root resorption.

It is interesting to note that the effectof the intra-cana! infection appearedmost pronounced in the margina! area.On!y 2 infected teeth displayed peri-apical patho!ogy. However, this is in ac-cordance with a previous investigationon the development of periapica!iesions in teeth with piaque infectedroot canals (Jansson et al. !993c) inwhich on!y minor periapica! granu-!omas had deve!oped after two months,whiie no inflammation cou!d be de-tected after one month. It thus, appearsthat spread of intra-canal bacteria ortheir toxins is more rapid in the cervica!area through patent dentina! tubu!escompared to the apica! area.

In conclusion, an intra-cana! infec-tion of endodontic pathogens stimu-lates epitheiia! downgrowth a!ong de-nuded dentin surfaees with margina!communication. Extrapo!ated to thechnica! situation, endodontic infectionsin periodontitis-prone patients mayaugment periodontitis propagation.

Acknowledgements

This study was supported by the Swed-ish Medica! Research Counci! (grantno. 665!), tiie Stoc!<:ho!m CountyCounci! (SLL) and the Facu!ty of Den-tistry at Karohnslia Institutet,

Zusammenfassung

Endodontische Pathogene und die Versidr-kung der ParodontalerkrankungVor kurzem wurde gezeigt, daB periapikalepathologische Prozesse, wenn sie bei Zahnenmit marginaler Parodontitis vorhanden sind,mit marginalem parodontalen Abbau korre-lieren. Dies wurde in Verbindung gebrachtmit offenen Dentin Tubuli im Zahnhaisbe-reich, einer Region, die nonnalerweise nacheiner Parodontalbehandlung frei von Zemeniist. Die Aussagekraft dieser Studien ist je-doch begrenzt, da die Kriterien ftir eine end-odontische Infektion nur ungenau definiertwurden, wie etwa Vorhandensein von peri-apikaler Destruktion, Das Ziel der vorliegen-den Untersuehung war es, an Wurzeloberfla-chen ohne Zement, jedoch mit gesundem pa-rodonialen Ligament die Wirkung vonendodontischen Pathogenen anf die margina-le parodontale Wundheilung zu messen, Hin-sichtlich der Ausbildung von pathologischemTaschen epithel und Bindegewebe wurden si-gnifikante Unterschiede zwischen infiziertenund nicht-infizierten Zahnen gefunden: Dieexpcrimcni-ellen Defekte wurden an den infi-

zierten Zahnen von ungefahr 2O'/o mehr Ta-schenepithel bedeckt, wahrend Defekte annicht-inflzierten Zahnen ungefahr 10% mehrBedeckung mit Bindegewebe zeigten. Darauswurde geschiossen, daO die intrakanalare In-fektion von endodontischen Pathogenen dasepitheiiale Tiefenwachstum entiang von ze-ment freien Dentinfiachen mit marginalerVerbindung stimuliert, Auf die klinische Si-tuation ubertragen, konnten endodontischeInfektionen bei Patienten mil Parodontitisdas Fortschreiten der Parodontitis erleich-tern.

Resume

Les pathogenes endodontiques dans I dugnien-tation de la maladie parodontaleUne correlation a recemment ete mise en evi-dence entre une destruction parodontalemarginale e: un etat patbologique periapicalindiquant une infection endodontique. lors-qu'il est present au niveau de dents atteintesd'une parodontopathie, Ce fait a ete associeavec la presence de canaiicules dentinairesouverts au collet de la dent, region normale-ment depourviie de cement apres le traiLe-ment parodontai Un inconvenient etait ce-pendant dans ces etudes !e fait que seules desprcuves indirectes, te!les que la presenced'une destruction periapicale, etaient utih-sees comme criteres de i'infection endodonti-que. Le present travail se proposait d'evaiucrles effets des pathogenes endodontiques surla guerison de la p!aie parodontale marginaleau niveau de surfaces radicuiaires depour-vues de cement, mais entourees d'un desmo-donte sain, Des differences significatives ontete trouvees entre les dents infectees et lesdents non infectees en ce qui concerne tes po-ches pathologiques et le tissu conjonctif, Leslesions experimentales etaient eouvertes d'en-viron 20% dc plus d'epitheiium de la pochedans ies dents infectees, tandis que les iesionsdes dents non infectees etaient eouvertesd'environ 10% de plus de tissu conjonctif. Enconclusion, une infection intra-canalaire pardes pathogenes endodontiques stimulo laproliferation epilheliale en dii^cction apicalele long des surfaces dentinaires denudees aveccommunication marginale. Si on appliqueces resultats a la situation clinique, les infec-tions endodontiques chez des patients predis-poses aux parodontopathies peuvent aug-menter la propagation de la parodontopa-thie.

References

Andreasen, J. O. & Andreasen, F. M, (1994)Avulsions. Chpt JO. In: Te.xtbook and col-our atlas of traumatic injuries to the teeth,eds, Andreasen, J. O. SL Andreasen, F. M.,3rd edition. Copenhagen: Munksgaard,1994, pp, 383-426.

Axelsson. P & Lindhe, J. (1981) Effecl ofcontrolled oral hygiene procedures oncaries and periodontal disease in aduits.Results after 6 yeai's. Journal of ClinicalPeriodontology 8, 239-248.

602 Jansson et al.

Berg. J.-O,, Blomlof, L. & Lindskog, S.(1990) Cellular reactions in pulpal andperiodonial tissues after periodonta]wound debridement. Journal of ClinicaiPeriodontohy 17, 165-173.

Blomiof, L,, Lengheden. A. & Lindskog, S-(1992) Endodontic infection and caieiumhydroxide-treatment. Effects on peri-odontal healing in mature and immaturereplanied monkey teeth. Journal of Clin-ical Periodontology 19, 652-658.

Blomlof, J,, Jansson, L,, Blomlof, L, & Lind-skog, S. (1995a) Etching at neutral pHpromotes periodontal healing. Journal ofClinical Periodontology TI, in press,

Blomiof, J., Jansson, L,, Blom]6f, L, & Lind-skog, S. (1995b) Long-time etching at lowpH jeopardizes periodonta! healing.Journai of Clinical Periodontology 22, 459-463.

Bowers, G., Chadroff, B,, Carnevale. R,,Mellonig, J,, Corio, R,, Emerson, J,, Slev-ens, M. & Romberg, E. (1989) Histologicevaluation of new attachment apparatusformation in humans. Part L Journal ofPeriodonlology 60, 664-674,

Caton, J. & Nyman, S. (1980) Histometricevaluation of periodontal surgery, L Themodified Widman flap procedure. Journalof Clinical Periodontology 7, 212-223,

Caton, 1, Nyman, S. & Zander, H. (1980)Histometric evaluation of periodonta]surgery, TL Connective tissue attachmentlevels after four regenerative procedures,Journai of Ciinical Periodontology 7. 224-231,

Cvek. M., Granath, L,-E, & Hollender. L,(1974) Treatment of non-vital permanentincisors with calcium hydroxide. Ill, Vari-ation of occurrence of ankylosis of reim-planted teeth with duration of extra-al-veolar period and storage environment,Odontoiogi.si< Revy 25, 43-56,

Ehnevid, H,, Jansson, L,, Lindskog, S, &Biomlof, L, (1993a) Periodontai healing inteeth with periapieal lesions, A clinicalretrospective sttidy. Journal of CiinicaiPeriodontoiogy 20, 254-258.

Ehnevid, H., Jansson, L., Lindskog, S. &Blomlof, L, (1993b) Periodontal healing in

pericdontitis-prone teeth in relation toradiographic attachment and endodonticinfection. Journal of Periodontoiogy 64,1199-1204.

Ehnevid, H., Jansson, L,, Lindskog, S,, Wein-traub, A. & Blomlof, L, (1994) Endodon-tic pathogens: propagation of infectionthrough patent dentinal tubules, EndodDent Traumutol 1994: in press,

Friskopp, J, & Blomlof, L. (1988) Spon-taneous periodontitis in a sample group ofthe monkey Maeaca Faseicularis. Journalof Periodonial Research 23, 265-267,

Hammarstrom, L., Blomlof, L,, Feigiin, B. &Lindskog, S, {1986} Effeet of eaieiuni hy-droxide treatment on periodontal repairand root resorption. Endod DentTraumawH, 184-189,

Jansson, L., Ehnevid, H., Lindskog, S, &Blomlof, L, (1993a) Relationship betweenperiapical and periodontal status, A clin-ieal retrospective study. Journal of CUnieaiPeriodontology 2^. 117-123,

Jansson, L,, Ehnevid, H,, Lindskog. S, &Blomlof, L, (1993b) Radiographie attach-ment in periodontitis-prone teeth with en-dodontic infection, Journai of Periodoitto-iogy 64, 947-953.

Jansson, L,, Ehnevid, H., Lindskog, S. &Blomlof, L, (1993c) Development of peri-apieal inflammatory lesions, Swedish Den-tal Journal 17. 85-93,

Jansson, L., Ehnevid, H,, Lindskog, S, &Blomlof, L, (1994) Proximal restorationsand periodontal status. Journai of ClinicaiPeriodontology 21, 577-582.

Karring.T., Nyman, S.. Lindhe, J. & Siriat,M, (1984) Potentials for root resorptionduring periodontal wound healing.Journai of Ciinical Periodontologv 11, 41-52.

Kerekes, K, & Olsen, L (1990) Similarities inthe mierofloras of root eanals and deepperiodontal pockets, Endod DentTraumatoi 6, 1-5,

Kipioti, A,, Nakou, M,. Legakis, N, &Milsis, F, (1984) Microbiological findingsof infected root canals and adjacent peri-odontal pockets in teeth with advancedperiodontitis. Oral Sitrgery 58, 213-220.

Knowles, J, W, Burgett, E G., Nissie, R. R.,Shiek, R. A., Morrison, E, C , Ramljord,S, P (1979) Results of periodontal treat-ment related to pocket depth and attach-ment level. Eight years. Journal of Period-ontology 50, 225-233.

Lehmann, E, L, (1975) Nonparametries: slat-isticai methods based on ranks, pp. 132-41.San Francisco: Holden-Day.

Lengheden, A. (1994) Periodontal impli-cations of caieium hydroxide treatment.Thesis, Stockholm: Karolinska Institute,

Lindskog, S, & Blomlof, L, (1994) fegenera-tion of the periodontal ligament. In: Theperiodontal ligament. 2nd edition, Mosby,London, in press.

Nyman, S., Lindhe, J. & Rosling, B, (1977)Periodonta! surgery in plaque-infecteddentitions, Jourtiai of Clinical Periodonto-logy 4, 240-249,

Nyman, S, & Karring, T. (1979) Regenera-tion of surgically removed buccal aiveoiarbone in dogs. Journal of Periodoniai Re-search 14, 86-92.

Pihlstrom, B,, McHugh, R., Oliphant, T,, Or^tiz-Campos, C, (1983) Comparison of sur-gical and nonsurgical treatment of peri-odontal disease, A review of currentstudies and additional results after 6 1/2years. Journal of Clinical Periodontoiogy10, 524-541,

Seymour, R, A,, Heasman, P A, (1992)Drugs, diseases, and the periodontium. Ox-ford University Press, Oxford,

Slots, J, & Taubman, A. (1992) Contempor-ary oral tnicrobiology and immunotogy.London: Mosby,

Sondqvist, G. (1976) Bacteriological studiesof neerotic dental pulps. Umea: UniversityOdontologicai Dissertations, no. 7, 1976,

Address:

E. JanssonDepartment of PeriodontologyPublic Denta! ServiceGotgatan 100S-I16 62 StockholmSweden