Ioannis K. KaioussisUrs BrdggerGiovanni E. ScdviWalter BuiginNiklaus P. Lang
Effect of implant design on survival andsuccess rates of titanium oral implants: aio-year prospective cohort study of theITI'^ Dental Implant System
Autbors' affitiation:loannis K. Kaioussis, Urs Briiggei. Giovanni E.Salvi, Walter BtiigiJi, Niklaus P. Lang, Departmentof Periodontology and Fixed Prostbedcs, Universityof Beme, Scbool of Dental Medicine, Beme,Switzerland
Correspondence to:Prof. Dr N.P LangUniversity of BemeSebool of Dental Medicine,Freiburgstrasse 7 CH 3010 BemeSwitzerlandTel.: +41 31 632 25 77Fax; 4-41 31 63J 49 15e-mail: nplang®dial.eunet.cb
Aim: The aim of this 10-year study (observation time 8-12 years, mean: 10 years) was to
compare the survival rates, success rates and incidences of biological complications using
three different implant designs of the ITI" Dental Implant System.
Material and methods: In 89 dental patients treated comprehensively, a total of 112 hollow
screw (HS), 49 hollow cylinder (HC) and 18 angulated hollow cylinder (AHC) implants were
installed depending on the available bone volume and according to prosthetic needs. One
and 10 years after surgical placement, clinical and radiographic parameters were assessed.
The incidences of peri-implantitis according to various thresholds were registered over 10
years of maintenance.
Results: Success criteria at 10 years were set at: pocket probing depth (PPD) s; 5 mm, bleeding
on probing (BoP)--, bone loss < 0.2 mm annually. The survival rate for HS was 95.4%, for HC
85.7% and for AHC 91.7%. Ninety percent of all the HS, 71 % of the HC and 88% of the AHC
did not present with an incidence of peri-implantitis over the 10 years, HC having
significantly higher incidence of peri-implantitis than HS (P< 0.004). With the success criteria
set above, a success rate for HSof 74%, for HCof 63% and for AHC of 6 1 % was identified at
10 years. However, including a definition of PPD^Gmm, BoP - and bone loss < 0.2mm
annually for success, the rates for HS were 78%, for HC 65% and for AHC 67%, respectively.
Basing success criteria purely on clinical parameters (without radiographic analysis), such as:
PPD^5 mm and BoP-, the success rates increased to 90%, 76% and 89%, respectively. With
PPD^6mm and BoP- as success criteria chosen, the respective rates were 94%, 82% and
94% for HS, HC and AHC implants, respectively.
Conclusions: A significantly higher survival rate as well as a significantly lower incidence of
peri-implantitis was identified for hollow screw design ITI" Dental Implants after 10 years of
service when compared to hollow cylinder design ITI " Dental Implants (95.4% vs. 85.7%;
10% vs. 29%). Depending on the setting of the threshold criteria for success, success rates
are highly variable and hence, reporting of success rates with elaboration on the criteria set
appears crucial for comparison of different studies.
Date:Accepted io fanuary 1003
To dte this article:Kaioussis (K, Bragger U, Salvi GE, BOrgin W, Lang NP.Effect of implant design on survival aiid succc^ rates oftitanium oral implants: a lo-year prospective cohortstudy of the m * Dental Implant System.Clin. Orai Impl. Res. 1$, 2txi4.j B-17
Karoussis et ai . Long-tenn impbnt prognosis of m'^ Denul Implants
These faetnrs may include smoking, un-controlled diabetes, colonization of theperi-implant sulcus with specific virulentmicnKirganisms.
Some systemic conditions such as osteo-porosis, radiation therapy, etc. have also|-)een imphcated (Genco et al. 1998). How-ever, the true significance and the clinicalimpact oi these factors for the longevity ofosseointegrated oral implants is - at thepresent time - still a matter of debate(Tonetti 1998).
It has been hypothesized that implantdesign may also play a role for thepreservation of jwri-implant marginal bone(Borchers ik Reichart 1983). The design of;m implant should allegedly not cause high-strain concentrations at the implant neckin order to avoid crestal bone resorption(Rieger et al. 19S9). It has been postulatedin in vivo studies (Pilliar et al. 1991) thatdifferent strain charaeteristies were mani-fested in the supporting bone aroundvarious implant designs.
Cylindrical, conieal, stepped, tapered,serew-shaped and hollow cylinder (HC|implant designs yielded variations in straindistribution in vitro m fitiite-element stu-dies [Siegele &. Soltesz 1989). Screw-shaped and full-bcxiy cylindrical implantspnxluced less strain than implants withsmall radii of curvature (eonieal), withgeometric discontinuities (stepped), or HCimplants. For cylindrical implants horizon-tal loading induced maximum strain in thecrestal aspects, while for screw-shaped im-plants maximum strain was ctineentratedin the regions helow the uppermost threads.
The m ^ Dental hnplant System hasevolved through several generations ofimplant designs originating from baskettypes and ranging from eccentrie to rotarysymmetrical HCs and hollow screws (HSs),the latest mcxst ctimmonly installed im-plant being the full-btxiy screw. Varyingsuccess rates have been reported for theprevious HC Type F implant (ten Bruggen-kate et al. 199OJ Versteegh et al. 1995;Mericske-Stem et al. 2001). Over 7 years offunction the novel ITI" Dental ImplantSystem with its hollow cylinders presenteda higher cumulative failure rate (8.7%)than the solid screw of the same ITI*Dental Implant System (3.2%). On theother hand, 4.2% of the HS implants werelost over 7 years in the same study (Buser etal. 1997).
Defining the clinical criteria for peri-implant conditions and hence, determiningcriteria for implant success has hitherto notbeen luiifonn. Even the terms survival,success, comphcation and failure rates havebeen employed with a wide range ofinterpretation {Sehnitman & Shulman1978; Albrektsson et al. 1986; Buser et al.1990; van Steenberghe 1997; van Steen-berghe et al. 1999I. However, for mostclinicians, implant failure represents thecomplete loss of the implant.
Nevertheless, implant failures have alsobeen defined as circumstances in which theconditions of the peri-implant tissues wereclinically not satisfactory (Mombelli 1994I.
Implant failures may better be addressedas biological comphcations which may ormay not lead to tbe loss of the implant.These may be classified according tochronological criteria as 'early' or 'late'.Early comphcations occur during the heal-ing period following implant installationand up to obtaining elinieally successfulstability, usually 6 months following in-stallarion. A variety of factors may beresponsible for early complications, suchas damage of the bone due to overbeatingduring prepararion of the recipient site, lackof primary stabihty of the implant, bacterialcontamination by improper surgical tech-nique, overloading by temporary recon-strucrion. Late complicarions, however,are observed after loading and are generallyassociated with the development of a peri-implant infeerion or peri-implantitis.
Previously, an implant system was ac-cepted to be 'successful' if the implants -on an average - lost not more than 1.5 mmof alveolar bone height in the first year offunction and not more than 0.2 mmannually in the subsequent years (Albrekts-son et al. 1986; Albrektsson & Isidor 1994).In more recent years, clinical parametershave been added to describe implant com-plications and included bleeding on probingof tbe peri-implant suleus and/or probingdepth exeeeding 5 mm (MombeUi & Lang
1994)-As defined at the ist European Work-
shop on Periodontology in 1993, peri-implant mucositis represents a 'reversibleinflammatory reaction in the soft tissuessurrounding a functioning implant' andperi-implantitis represents an 'inflamma-tory reaction with lass of supporting kmein the tissues surrounding a functioning
implant' (Albrektsson & Isidor 1994). It isevident that such definitions portray biolo-gical complications which may be treatedsuccessfully (Lang et al. 2000) in order toprevent total kws of the implant. If suchinterceptive treatment does not reach astatus of healthy peri-implant tissues theimplant may have to be explanted leadingto complete failure.
The aims of the present study were toassess and compare the io-year survivaland comphcation rates of HS, HC andangulated hollow cyhnder (AHCj m ' ^Dental Implants. To define criteria forsuccess of surviving implants, a hierarchyof various clinical threshold parameters wasapplied.
Material and methodsSubjects
Patients of a prospective, longitudinal,cohort study (Braver et al. 1997) of theDepartment of Periodontology and FixedPrasthodontics University of Beme Sehoolof Dental Medicine, were examined at iand 10 years after implant installarion. Thepatients had been treated for existingperiodontal disease according to a compre-bensive treatment strategy (Lang 1988)prior to the installation of implants andthe incorporation of suprastructures.
All the implants installed were implantsof the m * Dental Implant System (In-stitute Straumann AG, CH-4437, Walden-burg Switzerland). They were placedaccording to the manufacturer's guidelines(Sutter et al. 1988). The suprastructuresconsisted of single crowns or fixed partialdentures (FPD), which were seated between4 and 6 months postsurgjcally. The pa-tients were incorporated into the supportiveperiodontal therapy (SPT) program that wasprovided either at the Clinic for Period-ontology and Fixed Prosthodonties, Uni-versity of Beme, Switzerland, or in referringprivate dental practices at intervals between3 and 6 months.
At every recall examination during theio-year follow-up period, all biologicalcomphearions (peri-implanritis) were re-corded and treated according to the implantmaintenance and treatment protocol (cu-mulative interceptive supporrive therapy -CIST) (Lang et al. 2000).
9 I Clin. Oral Impl Res. 15, 2004 / 8-17
Karoussis et a l . Long-term implant progaosis of m ® Dental Implants
Clinical examination
Clinical i- and io-year evaluations in-
cluded the following parameters:
• modified plaque index (mPII) (Mombelli
etal. 1987I,• modified bleeding index (mBII) [Mom-
belli etal. 1987I,
• distance between the implant shoulderand the mucosal margin |DIM] in milli-meters [recession scored as negativevalue),
• pocket probing depth (PPD) in milli-meters,
• probing attachment level (PAL) in milli-meters, which was calculated by sub-tracting PPD from DIM,
• bleeding on probing (BoP).
Ail measurements were performed atfour aspects of eacb implant using a Hu-Friedy PGF-GFS periodontal probe (Hu-Friedy, Chicago, IL, USA). Readings weredone to tbe nearest millimeter.
Radiographic examination
Radiographs were obtained using a custo-mized Rinn filmholder (XCP" Instru-ments, Rinn Corporation Elgin, IL, USA)with a rigiti film-object-X-ray source beingcoupled to a beam-aiming device in order toachieve reproducible expostire geometry.The radiographs were captured using ahiack-and-white video camera (Canon,Still Video Products Group, Tokyo, Japan)and viewed on a light box. The images weretransferred to a personal computer (Com-paq 386/20, USA) and digitized with aframe grabber hardware card [Matrox Elec-tronic Systems MVP/AT, Dorval Quebec,Canada). Using an image-processing soft-ware, digitized images were stored with aresolution of 511 x 512 x 8 hit pixels (256shades of gray). Stored images were dis-played on a monitor and linear measure-ments were performed with the help of acursor (Bragger et al. 1996).
All radiograpbic measurements wereperformed in the 1 - and : o-year radiographsby one calibrated examiner (l.K.K.). Tbedistances in millimeters between tbeshoulder of tbe implant and the first clearhone to implant contact, mesially anddistally were noted. Changes in bone beightover tbe observation period as well asannual rates of change were calculated.
Success criteria for the implants at 10
years were set according to the following
hierarchy;
Success criteria I
1. Ahsence of mobility [Buser et al. 1990).
2. Absence of persistent subjective com-plaints (pain, foreign body sensationand/or dysaestbesia) (Buser etal. 1990).
3. No PPD>5 mm (PPD>s mm] (Mombel-li & Lang 1994; Bragger et al. 2001).
4. No PPD =5 mm (PPD =5 mm) andBoP + [Momhetli &. Lang 1994).
5. Absence of a continuous radiolucencyaround tbe implant (Buser ct al. 1990).
6. After the first year of service the annualvertical hone loss should not exceed0.2 mm [Alhrektsson et al. 1986; Al-brektsson &. Isidor 1994).
If a mesial or distal armual bone loss was>o.2mm, or a PPD [even at one implant-site) was >5mm, or PPD (even at oneimplant-site) was 5mm with BoP -!-, theimplant was characterized as 'unsuccessful'(implant witb a complication). 'Success'characterized an implant fulfilling Ixith theclinical and the radiograpbic criteria forsuccess mentioned [successful Implant).
Success criteria II (clinical success)
The criteria defined for this analysis werethe same as those specified alxjve [Successcriteria 1), but without including radio-graphic parameters.
Success criteria HI
Elevated threshold were defined as follows:
1. Ahsence of mohility [Buser et al. 1990).2. Ahsence of persistent subjective com-
plains [pain, foreign body sensation and/or dysaestbesia) (Buser et al. 1990).
3. No PPD >6nim (PPD>6mm) (Mom-helli &. Lang 1994; Bragger et al. 2001).
4. No PPD = 6 mm (PPD = 6 mm) andBop + (Mombelli & Lang 1994).
5. Absence of a continuous radiolucencyaround the implant (Buser et al. 1990).
6. After the first year of service, tbe annualvertical hone loss should not exceed0.2 mm (Albrektsson et al. 1986, Al-hrektsson & Isidor 1994).
'Success' tben characterized an implantpresenting no annual bone loss >o.2 mmmesially or distally, no site witbPPD>6 mm and no site with PPD ^ 6 mmand BoP-f (successful implant).
Success criteria IV (Clinical success)
Again, the criteria defined for this analysiswere the same as those specified above(Success criteria IE), but without includingradiographic parameters.
Statistical analysis
Survival rates
Descriptive statistics for all clinical andradiographic parameters were performedafter grouping tbe implants into HS, HCand AHC implants. For the comparisons ofthe clinical parameters and the bone losswithin the three groups, Kruskal-Wallistests were used.
For the estimation of survival rates andincidences of peri-implantitis, Kapl;in-Me-ier analyses were used (Kaplan & Meier1958). By means of life-table statistics, thecumulative survival rates were aUculated,using the following formula (van Steen-herghe et al. 1999):
CSR = {PCRS+{ISRxIOO-PCSR)}/^00
where i? —Race, C — Cummulative,
/ =Interval, P — Previous,
and S — Survival.
All implants that were not lost until theend of the observation period bave heenconsidered as censored according to the life-table statistics.
The sum of time periods of service in thedenominator is expressed as person-years,person-time or risk time. For each subject,the time at risk is the time during whichthis subject remains free of tlie disease, hutis at risk to develop disease (Ahlbom &NoreU 1992).
The definition mentioned may be trans-formed as follows:
Incidence of implant loss - number ofcases of implant loss in the study p<ipula-tion (13 implants) divided by the sum oflengths of time at risk for each implant. Allimplants were in danger either until theywere lost or until - more favorably - the endof the observation period. During theohservation period, all implant losses aswell as the exact time of service until thefailure was noticed were recorded. There-
1 0 I Clin. Oral Imp!. Res. i<i, 1004 / S-17
Raioussis et ai . Long-term implant prognosis of ITI* Dental Implants
fore, the survival time for each implantcould be accurately estimated, since bothimplantation and explantadon dates wereavailable. The incidences were calculated
as follows:
Incidence (I) forimplantioss
= number of losses/
y (time in service for each implant)
incidence of peri-implantitis during mainte-nance
Peri-implantitis was defined as an inci-dence of PPD^smni with BoP+ andratliographic signs of hone loss. For estima-tions of incidences of peri-implantitis,implants not affected until the end of theobservation period, were also considered ascensored. For the estimation of incidencesof pcriimplantitis the Kaplan-Meier analy-sis was used (Kaplan & Meier 1958). Afterstratification for implant types (HS, HC,AHC), the homogeneity of survival curves,I.e. differences hetween the groups for thesurvival rates and incidences of peri-im-[ilantitis, were tested hy log-rank test andWilcoxon test.
Success rates at 10 yearsx' and Fisher's exact test were used toevaluate differences in the success rateshetween the various groups of implant types.Morawer, frequency analyses were per-fonned for PPD around all implants at i;md 10 years of service. The evaluation ofsuccess rates included all 179 implantsinstalled at the lieginning of the stutiy (SiciliaFelechosaetal. 1999). The data analyses wereconducted hy means of the SAS statisticalsoftware (SAS Institute, Inc., 1999).
Of these, 112 were HS, 49 were HC and 18AHC m " Dental Implants.
Forty-Six of the 179 implants wereplaced in the anterior maxilla (10 HS, 26HC and 10 AHC), 5 in the anteriormandihle (2 HS and 3 HC| 58 in theposterior maxilla (40 HS, 14 HC and 4AHC), and 70 in the posterior mandible (60HS, 6 HC and 4 AHC).
At the io-year follow-up examination,the patient's ages ranged hetween 28 and 88years (mean age 58.9 years). The observa-tion period ranged from 8 to 12 years with amean of 10 years of service. Fiity-fiveimplants had been in function for 8-9years, 47 for 9-1 o years, s 3 for i o-i i yearsand 24 for 11-12 years.
The means for PPD, PAL, BoP, mPII,mBn, A Bone loss for mesial and distalaspects are presented in Tahle i. Significantdifferences were observed between HS andHC implants regarding mean PPD(1.60mm for HSy3.i4mm for HC,
P<o.oi8) and mean BoP (0.38 for HS/0.51for HC, pKO.oij,] as well as between HCand AHC with respect to mean mPII (0.42forHC/o.i8 for AHC, P<o.oi).
Survival and failure rates
A total of 13 of 179 implants were lost(failed implants) during the ohservationperiod resulting in a survival of 92.7% , oran incidence of implant loss of 7.3% in 10years.
With the Kaplan-Meier methtxl, thesurvival rates were properly estimated asheing 92.4% (SE: 0.02). For HSs, thecumulative survival rate was 95.4% (SE:0.02), for HCs 85.7% |SE: 0.05), and forangulated hollow cylinders (AHC) 91.7%(SE: 0.08) (Fig. i) (Table 2). Consequently,the failure rate of all implants was 7.6%after 10 years. For HSs it was 4.6%, forHCs 14.3%, and for AHCs 8.3% (Fig. i).
Table 1. Means, standard deviations (SD) and statistical test results (Kruskal-Wallis test) ofclinical and radiographic parameters measured for each implant
HS, 112 HC, n 49 AHC,n
mPII meanmBII meanPPD meanPAL meanBoP meanABone loss mesABone l05s dist
Mean
0.350.19
3.200.38""0.570.58
SD
0.410.280.890.890.341.171.04
Mean
0.42*0.193-143.590.510.770.84
SD
0.350.271.361.620.311.07112
Mean
0.180.213.123.430.461.461.71
SD
0.210.341.071.040.312.222.42
HS: hollow screw, HC; hollow cylinder, AHC: angulated hollow cylinder of the ITI" Dental ImplantSystem.
•Represents statistically significant difference compared to AHC (Kruskal-Wallis test: P<0.05)."Represents statistically significant difference compared to HC.
Results
Of the 127 patients examined at baseline, 917% (had passed away, 29 (23%) had moved,iway during the observation peritKi or werenot available for a complete re-evaluation.Hence, 89 patients (70%) of the originalcohort were available for the io-yeartollow-up examination. Of those, 34 weremale (38.2%) and 55 female {61.8%). Atthe time of implantation, the age of thepatients ranged from 19 to 78 years (meanage 49.33 years). A total number of 179implants were placed. Seventy-four of themwere placed in males and 105 in females.
100
40 60 80 100
months after implantation
120 140
Fig. I. The Kaplan-Meier estimate of survival rates oi ITI" Dental Implants as a function iif time sinceinstallation. HS: hollow screws, HC: hollow cylinders, AHC: angulated hollow cylinders.
I I I Clin. Oral Imp!. Res. 1$, 100^ I fi-17
Karoussis et al. Long-tenn implant piogQosis of m"^ Dental Implants
The homogeneity of the survival curveswas also tested (Altman 1991,- Hiisler &Zimmermaim 2001). Log-rank test showeda trend for significant {P<o.o6] differencesbetween HS and HC implants, with the HSto present a higher survival rate (95.4 vs.85.7). Wilcoxon test showed a statisticallysignificantly higher survival rate for HScompared to HC implants iP<o.O4).
In other words, the incidence (7) ofimplant loss (implant failure), calculatedby the Kaplan-Meier method, is:
/ —0.076 at 10 years or/ — 0.0076/iinplant-years or7=0.0076 implant losses per implant-
year or7.6 of 1000 m"^ implants are to be lost
per year.This evaluation reflects the mean risk for
the impiants of the study population to belost per year of observation (Ahlbom &.Norell 1992],
Similarly, the incidences of implant loss(implant failure) for HS, HC and AHCimplants were:
4.6 of 1000 HS type m " implants are tohe lost per year.
14.3 of 1000 HC type m * ' implants areto he lost per year.
8.3 of 1000 AHC type ITI* implants areto be lost per year.
Incidences of biological complicationsduring maintenance
84.6% (SE: 0.03) of all implants were free ofbiological complications throughout theobservation period. 90% (SE: 0.03] of theHS implants did not present hiologicalcomplications, while 71% (SE: o.oi) ofthe HCs and 88% (SE: 0.08) of the AHCremained free of any hiological comphca-tions (Fig. 2) (Table 3).
In order to test the homogeneity of theincidence curves, log-rank and Wilcoxontests were used. Both tests showed astatistically significantly lower incidenceof biological comphcations for HSs com-pared to HCs (log-rank test: P<o.ooo4 andWilcoxon test; P<o.oo3).
Success rates
Differences in the success rates betweenthe implant types were calculated using x̂and fisher's exact tests. With the first set ofsuccess criteria chosen (success criteria I), a
Table2. Survival and failure rates of ITI^ dental implants
Implanttype
Implantsplaced
Implantslost
Survival Failure Standarderror
HSHCAHCAll implants
11249^8
179
571
13
95.4*85.791.792.4
4.6'14.38.37.6
0.020.050.080.02
H5: hollow screw, HC: hollow cylinder, AHC: angulated hollow cylinder of the ITI* Dental ImplantSystem.•Represents 3 statistically significant difference compared to HC (Wilcoxon test, P<0.04).
100
20 40 60 80 100months after implantation
120
Fig. 2. The Kapian-Meier estimate of incidences of biological complications Iperi-implantitis) as a function oftime since implant installauon, HS: holtow screws, HC: hollow cylinders, AHC: an^ukted hollow cylinders.
Table3. Incidence of biological complications in ITI " dental implants
Implanttype
HSHCAHCAll implants
Implantsplaced
1124918
179
Implants withcomplications
1114
227
Complication-freeimplants (%)
90 '718884.6
Incidence ofcomplication (%)
10*291215,4
Standarderror
0.030.010.080.03
HS: hollov*/ screw, HC: hollow cylinder. AHC: angulated hollow cylinder of the ITI* Dental ImplantSystem.•Represents statistically significant difference compared to HC (Wilcoxon test, P<0.003, log-rank test.P<0.0004).
success rate of 74% for HS, of 63% for HCand of 61 % for AHC implants was found(Fig. 3 and Tahle 4). HS implants demon-strated a higher success rate when com-pared to cylinders. However, this differencewas not statistically significant.
Applying the second set of successcriteria (success criteria II), success ratesincreased to 90,2% for HS, 76% for HCand 86% for AHC implants, respectively.As shown in Table 4, HS implants pre-sented a significantly higher success ratecompared to HCs when relying only onclinical parameters (Fisher's exact test:P<O,O25).
If PPD^6mm in.stead o f ^ s m m waschosen as the definition of success (successcriteria HI), success rates of 78% for HS,65% for HC and 67% for AHC implantswere calculated (Table 5 and Fig. 3), Again,HS implants demonstrated a higher successrate compared to HC implants (78% vs.65%), However, this difference was notstatistically significant.
Relying on cUnical parameters only, hutwith the higher threshold for disease(success criteria IV: PPD<6mm andBoP-), the success rates increased to 94%for HS, 82% for HC and 94% for AHCimplants, respectively. HS implants pre-
12 I Clin. Oral Impl. Res. i$, 2004/8-17
KaiDussis et al . Long-temt implant prt^osis of ITI'^ Dental Implants
Tabte4. Success rates using 5mm as borderline for PPD and annual bone loss<0.2 mm.
Implant type ' Implants Success Clinicalplaced success'
HSHCAHCAll implants
1124918
179
83 (74%)31 (63%)11 (61 %)
125(69.8%)
101 (90,2%'l37 (76%)16 (89%)
154(86%)
HSi hollow screw, HC: hollow cylinder, AHC: angulated hollow cylinder of the ITI" Dental ImplantSystem,'Clinical success meaning that the implant fulfilled the clinical criteria,"Represents statistically significant difference compared to HC (Fisher's exact test, P<0.025).
Tables. Success rates using 6mm as borderline for PPD and annual bone loss<0.2mm.
Implant Type Implants Success Clinicalplaced Success*
HSHCAHCAll implants
1124918
179
87 (78%)32 (65%)12(67%)
131 (73.2%)
105 {94%'*)40 (82%)17(94%)
162 (85%)
HS: hollow screw, HC: hollow cylinder, AHC: angulated hollow cylinder of the ITI" Dental ImplantSystem.'Clinical success meaning that the implant fulfilled the clinical criteria."Represents statistically significant difference compared to HC (Fisher's exact test, P<0.024),
HS HC AHCB Clin. and Rx Success for PPD<5nim or PPD=5mm with BoP neg.
and annual bone ioss<0.2mm• Clin. Success (or PPD<5mm or PPD=5mm with BoP neg.
B Clin. and Rx Success tor PPD<6mm or PPD=6mm with BoP neg.and annual bone loss<0.2mm
• Clin. Success for PPD<6mm or PPD=6mm with BoP neg.
Pig. J. Success rates of m"^ Dental Implants for various thresholds of clinical and radiographic criteria for thedefinition of success.
sented a significantly higher success ratecompared to HC implants (Fisher's exacttest: P<o.o24).
Figure 4 reveals the cumtdative percen-tage of sites with various PPDs at i and 10years of service. The curve with the PPDvalues for the i o-year foUow-up was shiftedto the right compared to the i-year evalua-tion. The two lines at 5 and 6 mm define
the two thresholds set for the definition ofsuccess/complication criteria.
Discussion
In the present study, m ' * HS implantsdemtinstrated higher survival rates(95,4%), lower incidence of biological
comphcations (10%) and higher successrates compared to ITI" HC itnplants.Lower success rates for HC implants(91,3%) compared to HS implants(95.8%) had already been reported over anobservation period up to 7-years (Buser etal. 1997). In that study, however, cumtila-tive success rates for HS were similar to theones obtained with full screw m"^' im-plants (95,8% vs, 96.8%, respectively).The stirvival rates over 10 years of thepresent study are, therefore, in excellentagreement with those reported for up to 7years, althotigh a variety of stirgeons hadinstalled the ITT^ implants in the presentstudy as opposed to only one surgeon in theBuser et al. (1997) study. The possiblereasons for inferior survival rates of HCimplants in comparison to HS implants arelargely unknown and may only be specu-lated upon.
It is stirprisitig that no statisticallysignificant differences in survival rateswere found in the present study betweenHS and AHC implants. Lack of statisticalsigtiificance may be explained by the smallntunber of AHC (n=i8 l incorporated inthe present study. If nonaxial loading woiddhave compromised osseointegration, morefailures and complications with AHCwould have heen expected. In the contextof an argument on loading forces the restiltsof the present study to agree with thosereported by Bnlshi et al. (1997), whoquestioned the role of nonaxial loading asan etiologic factor for implant failure.
Using the definition of incidence forimplant loss, it can he concluded that only4.6 of 1000 HS implants, but 14.3 of 1000HC and 8.3 of 1000 AHC implants areexpected to be lost per year. Tbese expecta-tions appear quite satisfying, especially forthe HS implants and are in agreement withprevious reports on the same systerm[j-j-jH] [Busei- 1999; Buser et al. 1997,1999; Brocard et al. 2000; Hellem et al.2oot). Survival rates of implants fn)m thisand other implant systems were subjectedto a meta-analysis of 19 publications per-formed by Lindb et al. (1998). A total of2116 implants were included in this analy-sis. After 6-7 years of service, the survivalrate for implant-supporting fixed partialdentures |FPD) was 93.6%.
A distinction between different reasonsfor implant failtires is only made in a fewstudies. Usually, the number of implants
13 I Clin. Oral Imp!. Res. 15, 1004 / 8-17
Karoussis et al . Long-term implant prognosis of m * Dental Implants
100
80
60
40
20
11j
J
1—B D • •
- • - l O y r s
0 1 2 3 4 5 6 7 8 9 10 11
PPD in mm
fig. 4. Cumulative percentages of number of sites with respective mm of PPD at four sites of each impiant at iand 10 years after implant installation.
reported to he lost due to peri-implantitisappears low. On tbe basis of 73 selectedstudies, Esposito et al. (1998) publishedcombined data on early and late complica-tions of Branemark implants. Over a 5-yearperiod, a biologically related failure rate(implaiit loss) of 7.7% was noted.
Ellegaard et al. [1997) evaluated thesuccess of implant therapy in periodontallycompromised patients. After s years, 70%of the m implants were free of pockets>6mm, while after 3 and i year 92.3%and 96.3% of the implants had heen free ofsucb pockets (S 6 mm). Setting tbe thresh-old for the definition of a biologicalcomplication at PPD ̂ s mm and BoP + ,Bragger et al. (20Q1) found a prevalence ofperi-implantitis of 9.6% at implants exam-ined after 8 years. With a threshold of^ 6 mm and BoP + as the definition forperi-implantitis, tbe prevalence was re-duced to 5 % at 8 years.
During the present study, all incidencesof biological complications (peri-implanti-tis) were noted. 15.4% of all implantsplaced presented a complication duringthe 10 years of observation. Complicationsoccurred at 10% of hollow screw implants,while at HC implants, the prevalence ofperi-implantitis in lo years was almostthree times higher (29%). AHCs presenteda complication rate of 12% (Fig. 2) [Table3). HSs, therefore, presented a lower chanceto be affected hy biological complications.
Success rates depended on tbe thresholdscbosen to distinguish hetween success and
complications (Fig. 3). Tbe diagnosticinformation to set different thresholdsincluded the assessments of BoP, PPD andcbange in radiograpbic bone height. Therationale for selecting these criteria asrelevant parameters for continuous evaltia-tion ratber than only survival rates, lieswitbin tbe fact that hiological complica-tions may or may not lead to the loss of animplant: Provided tbat the diagnosis of thecomplication is made soon after the hiolo-gical onset of tbe event, the complicationmay he dealt witb early enougb in order toavoid further progression of the lesion.
After I year of function, the frequency ofBop was twice as bigh at implant sitescompared to matching control tootb sites ina group of patients similar to those incor-porated in the present study [Bragger et al.1997). Periodontal sites reveahng repeatedBoP were at bigber risk to loose attach-ment, whereas the absence of BoP docu-mented high negative predictive values forteeth (Lang et al, 1994). Similar diagnosticinput may he expected from tbe evaluationof BoP at implant sites [Luterhacber et al.2000).
PPD reflects tbe amount of tissue resis-tance to probing. Deptb force pattemsrevealed cbaracteristic differences at teethcompared to implants and depending on theprobing forces apphed [van der Velden et al.1979; MombeUi 6L Graf 1986; MombeUi etal. 1997). Peri-implant sites with increasedPPD were associated witb a patbogenicmicroflora (Kalyakakis et al. 1994). These
sites may reflect a condition of peri-implantitis recommended to be treatedaccording to a CIST protocol (MombeUi &Lang 1998). In histologic samples, Lang etal, [1994) documented tbat tbe measure-ment error of PAL readings depended on theperi-implant healtb conditions. It is tlierc-forc important to assess both inflammatoryand pocket probing status. Both 5 mm and6 mm PPD were therefore chosen to setthresholds. Either as clinical parameteralotie or in combination witb radiographi-caUy assessed loss of peri-implant bone.
With the success criteria set, a successrate of 74% for HS, of 63% for HC and of61% for AHC was found [Table 4). This isin contrast to the findings of Buser et al.(1997), who found a cumulative successrate of 95.8% at 7 years for HS and of91.3% for HC. Brocard et al. [2000),however, reported on a cumulative successrate of 83.4% at 7 years. Buser etal. (1999)presented a cumulative success rate reach-ing 91.4% at i - io years, while HcUem etal. [2001) noted a success rate of 91.4%after 5 years.
Differences in success rates were theresult of the different criteria used for thedefinition of success. As an example fromthe hterature, HeUem et al. [2001) consid-ered an annual hone loss 0,5 mm asacceptable, while studies that used thecriteria of success for m implants assuggested hy Buser et al, (1990) did notinclude any PPD measurements. The pre-sent study utilized strict and objectivecriteria, i.e. PPD^5mm and BoP-.Furthermore measurements of peri-implantcrestal hone loss were included in theevaluation [bone loss <o.2 mm annually).If one site of an implant did not fulfiU thepredefined criteria, the implant was char-acterized as nonsuccessful (implant witb acomplication). Another possible infltienceon variations within the success rates ofsimilar studies may be attributable tovarious patient pools being evalated. Ahistory of periodontal disease in somepatient groups may indeed affect the in-cidence of peri-implantitis as demonstratedrecently (Karoussis et al. 2003).
Owing to the lack of standardized andinternationally recogtiized criteria for tbedefinition of success [van Steenhergheetal.1999), tbresbolds for peri-implantitis andother biological complications will con-tinue to be disctissed. In the present study.
14 ' Clin. Oral Imp!. Res. 15, 2004 / 8-17
Karoussis et a l . Long-term implant prognosis of m * ' Dental Implants
completely different success rates wereobtained depending on the criteria chosen.
The choice of PPD < 6 mm instead ofaffected the success rates
. Furthermore, relying on purelyclinical instead of a combination of clinicaland radiographic parameters, the successrates further increased. Setting of thresh-olds for success criteria is crucial forreporting success rates or incidences ofbiological complications in longitudinalstudies. In the present study, HSs demon-strated higher success rates compared toHC implants at any threshold chosen.
Conclusions
Within the limits of this prospective cohortstudy, the following conclusions are to bemade:
1. A lO-yearstuvival rate of 95.4% for HS,of 85,7% for HC, and of 91.7% for AHCm " Dental Implants was found.
2. HS presented a significantly higher sur-vival rate, lower incidence of biologicalcomplications, and higher success rates,compared to hollow cyhnder implants.
Acknowledgements: This study wassupported by the Clinical ResearchFoundation (CRF) for the Promotion ofOral Health, University of Berne, BerneSwitzerland and by the PapavramidesFoundation, Universities of Athens,Greece and Berne, Switzerland. Thesenior author is the recipient of an ITI-Scholarship for 2001-2002.
Resume
Lc but lie ce suivi li'unc decennic a Otc dt comparerles taux de survie, les taiix de aucccs et les incidencesties complications biologlques de l'utilisation de trois[iiodclcs implantaires differents du systeme ITI '̂'Denwl Implant. Chez Sg patients, 111 vis creuses|HS|, 49 cylindies creux |HC| et 18 cytindns crcuxangiilcs (AHC) ont ete places suivant !e volumeiisst'ux disponible et les necessites protlietiques. Uneui dix ;innecs aprte leur plaeemeni. des parametresi:liniques el radiii^raphiques ont ete definis, Lesincidences de paroiniplantitc relatives aux differentssculls ont ete enregistrees durant ces dix annees demaintenance. Les succes des chtcres a dix ans ctaientplaces a : PPD^s mm, BoP-, perte osseuse <o,2mm/an. Le taux dc suivie pour HS ctait de 9s,4%,|x)ur HC de 86% et pour AHC de 92%. Nonante(Xjur cent dc tous les HS, 71% des HC et 88% desAHC ne prcsentaient pas d'incidences de paroim-plantite durant ces dix annees, HC ayant une plus
importante incidence de paroifmplantitc que HS(p<o,oo4). Grace aux criteres de succes indiques,un taux de succes de 74% pour HS, de 63% pour HCct de 61 ?'o pour AHC a etc identifie apres dix ans .Cependant, en dcHnissant le succes avecPPD ^ 6mm, BoP- et pcrtc osseuse <o,2mm/an, lestaux ctaient de 787u pour HS, de 65 % pour HC et de67% f>our AHC. En basant les taux de sucetsuniquement sur les parametres ctiniques (sans['analyse radiograpliiqucl tels que PPD^5 mm etBoP-, les taux de succes augmentaient respective'mem a 90, 76 et 89%. Avec PPD^Sfemm et BoP-comme criteres de succes, les taux s'clevaientrespcctivcment a 94, 8a et 94%, Un taux de surviesigniticativemcnt plus important ainsi qu'une in-cidence significativement plus faible de paroimplan-tite etaient constates au niveau des implants viscreuses apH:s dix ans de mise en fonction comparesaux cylindres creux, Suivant l'etablissement du seuilpour les criteres du succcs, les uux de ce succes sontextrernement variables et rapporter les taux dc successuivant I'elahoration des criteres est done crucialpour comparer differentes etudes.
Zusammenfassung
Der Einflass des Implantatdesigns auf die Oberle-bens- und Erfolgsrate von Titanimplantaten: EineLangzeitstudie des ITr^'-Systems ufaer /o lahre.
Ziel: Das Ziel dieser io-Jahresstudie IBeobachtungs-zeit 8-12 lahre, Mittelwert: 10 lahre) war es, beim^-Implantaten mit drei verschiedenen Designs,die Erfolgs- und Oberlebensrate zu vergleichen, undtlas Auftrcten vou biologischen Zwischenfallen zuunteisuchen.
Material und Methiide: Bei 89 synoptisch behandel-ten Patienten implanticrtc man in Abhangigkeit desvorbandenen Knochenvtilumens und der pnithe-tischen Anfordemngen insgesamt 11 j Hoblschrau-benimplantate (HS], 49 HohlzylinderimpLmtate(HC) und 18 abgewinkelte Hohlzylindenmplantate(AHC). Ein und zehn Jahre nach der Implantationnahm man die klLnischen und radiologisehcn Para-meter auf. Eine Periijnplantitis registricrte wahrendder lo-jahrigen Eriialtungspliase anhand verschied-encr C.renzwerte.
Resultate: Die Kriterien fiir einen Erfolg nach 10lahren legtc man bei den folgenden Werten fest: PPD<5mm, BOP-, jahrlicher Knochenverlust <o.2mm.Die Oberlebensrate far ein HS lag bei 95.4%, filr einHC bei 85.7% und fur ein AHC bei 91.7%. 90%aller HS, 71 % aller HC und sa % aller AHC zeigtewahrend den lo Jahren nie Anzeicben einer Periim-plantitis, wohei die HC signiBkant haufiger Periim-plantitis hatten. als die HS (p«:o.oo4). Mit den obenfestgelegten Erfolgskriterien crgab sich nach 10Jahren for die HS eine Erfolgsrate von 74%, fflr dieHC eine von 63% und fur die AHC eine von 61%.Veranderte man die Definition auf "PPD<6mm,BOP -, jahrlicher Kochenverlust <o.2mm", sobetrugen die Erfolgsraten ftlr die HS 7R%, fur dieHC 6s% und fOr die AHC 67%. Basierten dieErfolgskriterien rein auf kiinischen Parametem |PPD<smm, BOP-, keine rontgcnologische Aniyse), sostiegen die Erfolgsraten auf 90%, 76% und 89% an.Wahlte man die Erfolgskriterien "PPD temm und
BOP •", so betrugen die Erfoigaraten filr die HS 94%,fQrdieHC8i% und fdr die AHC 94%.Zusammenfassung: Verglich man nach lo-jahrigerFunktion die Hiihlsciiraubcn des ITI"-Implantat-Systems mit den Hohlzylindeni dessclbcn Systems,so ergah sich filr die HS sowohl eine signiflkanthflher Oberlebensrate, wie auch ein seltencresAuftretcn von Penimplantitis 195.4% vs. 85.7%)10% vs. 29%). Die Erfolgsraten variieren in Abhdn-gigkeit der in der Definition eines Erfolgcs festge-legten Grenzwerten enomi stark. Dies erweist sichim Vergleich von verschiedenen Studien als hinder-lich, weil die Definition der Erfolgsraten meist aufverschiedenen Kiiterien beruhen.
Resumen
Intendon: La intencion de este estudio de ro aflos(tiempo de observacion 8-12, media: 10 aftos) fuecomparar las indices de supervivencia, intliees deexito e indices de complicaciones biologicas usandotres diferentes diacfios de impkntes del Sistema deImplantes Dentales m"^.
Material y Metodos: Se instalaion en 89 pacientesdentales tratados completamentc un total de t ntomill(»s huccos (HS], 49 cilindros huecas (HC) y 18cilintiros huccos anguladris (AHC) dependicndo de ladisponibiLdad de voUuiicn osco y df acucrdo con lasnecesidades protesicas, Sc vaK>ramn parametrosclinicos y radiografictis uno y <iiez afios trati lacolocacinn quinxi;gica. Se registramn Us incidenciasde periimplantitis de acuerdo con varios umbrales alo largo de to anos de mantenimiento.Resultados: Los criterios de exito a los 10 aAos sesituaron en PPD^^mm, BoP-, perdida (isea <0,1mm por ano. El indice de supervivencia para losHS fue del 95-4%, para los HC del 8 s,7% y para ioaAHC del 91-7%. El 90% de los HS, el 71% de losHC y el 88% de los AHC no presentanin ningunaincidencia de periimplantitis a lo largo de los r o aAos,HC tuvo una significativamente mayor incidenciade periimplantitis que HS (p< 0.004). Con loscriterios dc exito antes mencionados, se identifierun indice de exito para HS del 74%, para HC del6}% y para AHC de! 61% a los 10 artos.De t(xias modtKS, incluyendo una definieion dcPPD^6mm, BoP- y pcrdida osea <o.2mm al aflopara tener exito, los indices para HS fueron del 78%,para HC 65% y para AHC 67%, respectivamente.Basando los criterios de exito puramente en para-metros clinicos (sin analisis radiograflcos), talescomo: PPD^^mm y BoP-, los indices de ^ t osubieron hasta el 90%, 76% y 89%, respcctiva-metite. Con cl PPD ^ 6mm y BoP - como criterios deexito elegidos, los indices respectivos fueron del94%,82% y 94% para implantea HS, HC y AHC,respectivamente.
Condusiones: Se identified un significativamentemayor indice de superviveneia al igual que unamenor incidencia de pedimpiantitis para et disefto detomillo hueco Implante Dental m " . (9S.4% vs.8s.7; 10% vs. 29%). Dependiendo de la definieiondel criterio del umbral de exito, los indices de exitoson altamente variables y ]X)r tanto, los informcs delos indices de exito con elaboracion de la definieionde criterios parece scr crucial para la comparacion delos diferentes estudios.
15 I Clin. Oial Impl. Res. is, J004 / 8-17
Karoussis et a l . Long-term implant prognosis of m ^ Dental Implants
teS : 1 0 ^g^miS-aiJK^Ii : P P D:<.5 mm,B o P - , ip^-t!lf5';*< 0. 2mmTfc':>fc.t 1?^* )±HS95 . 4%, H C 8 5 . 7%, AHC9 1- 7 % - C t . o t . •^T:mHSm90%. HC!D71%, AHCC0 8 8%(i 1 O^rai;t>ft:»)
u s iztt-^-i y« A > o t ( p < 0 . 0 0 4 ) .
-iv^fc 1 O ^ » » ^ ^ * « H S 7 4 % , H C 6
3 % , A H C 6 I %-C*)oft;. LTJ'L' ' .£*! fj h£5ft
. B o P-AtT^Rfl-t«9^t , ^5 ! i^ ( i#* HA 7 8%,
HC65%, AHC67%T*-ofc. P P D <.5mmt B o P -
, 7 6 % t 8 9%-Cfcofc^ P P Do P-*ia5!)ftflS,fe4%, HC82%.
1>« ITI® -i y/y y Vit,
(95
If fjo$ j);
7%,
References
Adell, R., Eriksson, B., Lekholm, U., Branemark, P.& Jemt, T. (1990] Along-term follow-up study ofosseointegrated implants in the treatment oftotdly edentulous jaws. International loumal ofOzal and MaxUlofacial Implants 5: 347-3S9.
Ahlbom. A. & NoreU. S. I1991I Introduction toModem Epidemiology, 2nd edition, 26-31.Athens: Litsas Medical Editions.
Albrektsson, T. & Isidor, F. (1994) Consensus reportof session IV. In: Lang, N.P. & Kamng, T., eds.Proceedings of the 1st European Workshop onPeriodontology, 365-369. London: QuintessencePuhlishing Co., Ltd.
Albrektsson, T., Zarii, G., Worthington, P. &Eriksson, A.R. (1986) Tbe long-term efficacy ofcurrently used dental implants. A review andproposed criteria of success. Intemational jour-nal of Oral and Maxillofacial Implants i:ii-iS-
Balshi, T.J., Ekfeldt, A., Stenberg, T. & Vriehnck, L.I1997) Three-year evaluation of Branemark im-plants connected to angulated abutments. Inter-national foumal of Oral and MaxillofacialImplants 12: 52-58.
Borchers, L, & Reichart, P. (r983] Three dimen-sional stress distribution around a dental implantat different stages of interface development.foumal of Dental Research 62: 155-159.
Braver, U., Aeschllmann, S., Biirgin, W., Ham-merle, H.F.C. & Lang, N.P. (2001) Biological andtechnical complicariotis and failures with fixedpartial dentures |FPD) on implants and teeth afterfour to five years of funcrion. Cliiucal OralImplants Research 12: 26-34.
Bragger, U., BOrgin, W., Hammerle, C.H.F. & Lang,N.P. (1997) Associarions between clinical para-meters assessed around implants and teeth.Clinical Oral Implants Research 8: 412-421.
Bragger, U., Hugel-Pisoni, C, Biirgin, W., Buser, D.& Lang, N.P. (1996) Correlations betwt;en radio-graphic, clinical and mobility parameters afterloading of oral implants witb Bxed partial den-tures. A 2-year longitudinal study. Clinical OralImplants Research 7: 230-239.
Brocard, D., Barthet, P., Baysse, E., IXiifort, |-F.,Eller, P., Justumus, P., Marin, P,, Oscaby, F.,Simonet, T,, Benque, E. & Brunei, G, (2000) A
multicentre report on 1022 consecutively placedm implants: a 7-year longitudinal study. Inter-national loumal of Oral and Maxillofacial Im-plants 15: 691-700.
ten Bruggenkate, CM., Muller, K. & Oosterbeek,H.S. 11990) Clinical evaluation of the ITI(F-Type) hollow cylinder implant. Oral Surgery.Oral Medicine, and Oral Pathology 70:693-697.
Buser, D. (1999) Effects of various titanium surfaceconfigurations on osseointegration and clinicalimplant stability. In: Lang, N.P., Karing, T. &Lindhe, J. eds. Proceedings of the 3rd EuropeanViforkshop on Periodontology, 88-iai. Berlin:Quintessence Publishing Co., Ltd,
Buser, D., Mericske-Stem, R., Bemard, ],P,, Beh-neke, A., Behneke, N., Hirt, H.P., Belser, U.C, &Lang, N.P. (1997) Long-term evaluation of non-submerged m implants. Part i:8-year life tableanalysis of prospective multi-center study with2359 implants. Clinical Oral Implants Research8: 161-172.
Buser, D., Mericske-Stem, R., Dula, K. & Lang,N.P. I1999I Clinical experience with one-stage,non-submerged dental implants. Advances inDental Research ty, i53-i6t,
Buser, .D., Weber, H.P. & Lang N, P. (1990)Tissue integration of non-submerged implants,i-year results of a prospective study with rtxjm hollow-cylinder and hollow- screwimplants. Clinical Oral Implants Research i:33-40.
EUegaanl, B., Baelum, B. & Karring, T. (1997)Implant therapy in periodontally compromisedpatients. Clinical Oral Implants Research 8:180-188.
Ericsson, I,, Persson, L.G., Berglundh, T., Edlund, T.& Lindhe, ), I1996) The effect of antimicrobialtherapy on periimplantitis lesions. An experimen-tal study in the dog. Chnical Oral ImplantsResearch 7: 320-328.
Esposito, M,, Hirsch, J.M., Lekbohn, U, & Thomp-son, P. (1998) Biological factors contributing tofailures of osseointegrated oral implants. |I) Suc-cess criteria and epidemiology. European foumalofOralScienc.es 106: 527-551.
Genco, R.I., Gluridi, L, Haraszthy, V,, Zambon, J.& DeNardin, E. {1998) Overview of risk factors forpedodontal disease and implications for diabetes
and cardiovascular disease. Compendium of Con-tinuing Education in Dentistry 10: 40-45.
Hellem, S,, Karlsson, U., Bmnell, G., Hamp, S.E. &Astrand, P, [2001] Nonsubmerged implants in thetreatment of the edentulous lower jaw: a 5-yearprospecbvc longitudinal study of ITI holliiwscrews. Clinical Implant Dentistry and RelatedResearch 3: 20-29.
Htisler, I. & Zimmermann, H. (2001] StatistischePrinzipien far medizinische Projekte. 3 Auflage,218-225. Bem: Verlag Hans Huher.
Isidor, F. |r996] Loss of osseointegration caused byocclusal load of oral implants. A chnical andradiographie study in monkeys, Chnical OralImplants Research 7: 143-152.
Kalyakakis, G., Zafiropoulos, C.G., Muiat, Y,,Spiekemiann, H. & Niesengard, R,). (1994)Clinicat and microbiological status of osseointe-grated implants, foumal of Periodontology 65:766-770.
Kaplan, E. & Meier, P. (1958] Nonparametric esti-mation from incomplete observations, foumal ofthe American Statistical Association j j : 457-481.
Karoussis, I.K,, Salvi, G.E,, Mayfield, L,).A,, Brag-ger, U., Hammerle, G.H.F, & Lang, N.P. (2003)Long-term implant prognosis in patients with andwithout a history of chronic periodontids - a 10-year prospective cohort study of the m ^ DentalImplant System. Clinical Oral Implants Research14: 329-339.
Lang, N.P. I1988I Checklisten der Zahnmedizin.Stuttgart, New York: Georg Thieme Verlag,
Lang, N.P., Joss, A,, Orsanic, T,, Gusberti, F.A. ft.Siesrist,B.E. 11986) Bleeding on probing. A predic-tor for the progression of periodontai disease?loumal of Clinical Periodontology 1%: 59c>-596.
Mombelli, A. I1994I Criteria for success. Monitoring.In: Lang, N.P. & Karring, T., eds. Proceedings of theIst Euntpean Wf>rkshop on Penodontology, 317-325. London: Quintessence Publishing Go., Ltd.
Mombelli, A. & Graf, H. I1986) Depth-forcepatterns in periodontat probinig. foumal of ClinicalPerio<lontoh}gy i j : 126-130.
Mombelli, A, & Lang, N.P. (1994) Glinical para-meters for the evaluation of dental implants.1'n.riodontolc^ 2000 4: 81-86,
Mombelli, A. & Lang, N,P. I1998) The diagnosisand treatment of periimplantitis. Pmodontology2000 17; 63-76.
Mombelli, A., Muhle, T., Bragger, U., Lang, N,P, &Buigin, W. 11997) Comparison of periodontal and|.X'ri-implant probing by depth-force pattem analy-sis. Clinical Oral Implants Research 8: 448-454.
Momhelli, A., Van Oostcn, M.A,G., Scbttrch, E. &Lang, N.P. |i987)Theiiiicrobintaa.ssociatedwithsuccessful or (ailing osseointep'ated titaniumimplants. Oral Microbiology and Immunology2:145-151
Pereson, L.G., Ericsson, L, Beiglundh, T. & Lindhe,J, 11996) Guided bone regenerarion in the treat-
ment of periimplantitis. Clinical Oral ImplantsResearch 7: 366-372,
Pilliar, R.M., Deportet, D.A., Watson, P.A. &Valiquette, N, {1991! Dental implant design effecton bone remodeling, loumal of BiowaterialsResearch x$: 467-483,
Rieger, M,R., Fareed, K., Adams, W.K. & Tanquist,R.A. |t989l Bone stress distribution for threeendosseous implants, joumal of Prosthetic Den-tistry 61-. 123-228.
Rosenberg, E,S,, Torosian, J.P. & Slots, |. (1991!Microbial differences in 2 clinically distinct typesof failures of osseointegrated implants. ClinicalOral Implants Research 2: 135-144,
Rutar, A,, Lang, N.P., Buser, D., Btirgin, W, &Mombelli, A. (2oor) Retrospective assessment dclinical and microbiolc^cal factors affecting peri-implant tissue contlitions. Clinical Oral ImplantsResearch 12: i89-r95,
SAS Institute Inc, (1999) SAS system. Version 8.0.Cary, NC: SAS Institute Inc.
Schnitman, P.A. & Shulman, L,B. I1978I Recom-mendations of the consensus development am-ference on dental implants, loumal of theAmerican Dental Association 98: 373-377.
Sicilia Felechosa, A., Arregui Hurtado, I. & Men^-dezGollar, M. (1999) Requirements for the designof clinical trials in implant dentistry. In: Lang,N.P., Karring, T. & Lindlie, [., eds. Proceedings ofthe 3rd European Workshop on Periodontology.Implant Dentistry, 194-216. Berlin: Quintes-sence Publishing Co., Ltd.
Siegelc, D. & Soltesz, U. (1989) Numerical inves-tigations of the influence of implant shape onstress distribudon in the jaw bone. Intemational
Jcaimal of Oral and Maxillofacial Implants 4:333-510.
van Steenberghe, D. (r997) Outcomes andtheir measurement in clinical trials of endosseousoral implants, A nnak of Periodontology a:291-298,
van Steenbeighe, D,, Qtiirynen, M. & Naen, I,[1999) Stuvival and success rates with oralendosseous implants, in: Latig, N,P., Karring, T,61 Lindhe, ]., eds. Proceedings of the 3nl EuropeanWorh'ihop on Periodontohgy, Implant Dentistry,142-254, Berlin: Qtiintessence Publishing Co.,Ltd,
Sutter, F,, Schroeder, A, iii Buser, D. 119SSI The newconcept of m hallow-cyUnder and hollow-screw implants: Part 2. Clinical aspects, indica-tions, and early clinical results. Intemationalloumal Oral and Maxillofacial Implants 3:173-181.
