Clinical XP

clinical Experience WithEmpress Crowns

Mauro Fradeani, MD, DDSSpecial Lecturer, Department of ProitbodonticsLouisiana State University:and Private PracticeFesaro and Milano, Italy

Augusto Aquilano, DDSPrivate PracticePesaro, Italy

The IPS Empress pressed glass-ceramic system was used in this investigationto restore anterior and posterior single teeth. One hundred forty-four crownswere evaluated over a period of 6 to 68 months (mean period of 37 months).According to the Kaplan-Meier analysis, the estimated success rate after thisperiod was 95,35%, Crowns were also investigated using the modified USPublic Health Service critena. Most of the crowns rated as Alpha for colormatch, contour, marginal integrity, recurrent caries, and marginal discolor-ation. The results of this study indicate that this material can be successfuiiyused, especially in the anterior area, when the procedures outlined drecarefully followed, Int j Prosthodont l997;10:24l-247.

The esthetic treatment of teeth involving the useof single complete-coverage restorations offers

several options. The fraditional mefal ceramic sys-tem allows maximum strength, but at the sametime the light transmission properties are com-pletely differenf from those of nafural teeth be-cause of the mefal framework,' The increased es-thetic demands of the patient can be satisfied byusing one of fhe numerous all-ceramic materialsdeveloped in recent years.

To maintain the same strength obtained by metalceramic restorations, some all-ceramic systemssuch as the In-Ceram system (Vita, Zahnfabrik, Bad-Säckingen, Germany) use a reinforcing aluminousoxide ceramic core,^'^ The reduced size of fhe alu-mina crystals compared to that of previous alumi-nous sysfems gives befter esthetic results, eventhough the core is still relatively opaque. Wheregreater translucency is desirable, the In-CeramSpinell system offers a substantial improvement,""

Of the ceramic systems available, however,glass-ceramic materials such as Dicor (Dentsply,York, PA),̂ '6 the I eue i te-re info reed glass-ceramicOptec HSP (American Thermocraft, Somerset, NJ),''and the leucite-reinforced pressed glass-ceramic

Reprint Requests: Dr Mauro Fradeani, Corso XI Settembre, 92.

6!mOPesaro, Italy.

IPS Empress (Ivoclar, Amherst, NY) can ensure anexcellenf gradienf of translucency, very similar tothat of natural dentition^ (Figs 1 and 2), TheFmpress system offers substantial advantages withnondiscolored teeth, including improved strength.

IPS Empress is a heat-pressed ceramic: the glass-ceramic ingot is partially precerammed by the man-ufacturer and then processed in the laboratory andcompleted using either a surface-coloration or lay-ering technique. The latter is primarily used to fabri-cate anterior crowns, while the surface-colorationtechnique is adopfed for posterior restorations, assuggested by the manufacturer.

After wax elimination, the glass-ceramic ispressed into the preheated muffle, A femperatureof 1,050''C is required for pressing the completeform, and a femperature of l,180°C is required iffhe layering technique is to be used. Both tech-niques require .5-bar pressure for 30 to 40 minutes.When the layering technique is used, the substruc-ture is covered with dentin and enamel porcelainfollowed by a final glazing cycle. For the surface-coloration technique, the restoration form is ob-tained directly from fhe wax-pattern,

A study by Dong et al^ reported a significanf in-crease in the flexural strength of the precerammedingot (74 MPa) after a heat-pressing treatment (126MPa), Subsequent heat treatments increase the IPS

n o . Numbers, 1997 241 The International Journal of Prosthodontii

Ciinic.ii Eïperienœ Wilh Empreîs Ciow Frsdeani/Aqu

Fig 1 Preoperative view cf maxillary central incisors withfaiied acryiic resin crowns.

Fig 2 The proper gradient of transiucency ailows a good in-tegration of the two Empress restorations.

Tabie 1 Distribution ot Restorations by Tooth Type

MaxiilarvMandibularTotal

Centralincisors

474

51

Lateralincisors

335

38

Canines

93

12

Premolars

101828

Firstmclars

47

11

Secondmoiars

134

Total

10440

144

Table 2Position

Distribution ot Restorations by Tooth

MaxillaryMandibuiarTota i

12101

10440

144

Empress strength to 182 MPa. Surface glaze firingsfurther raise the material strength to 215 MPa.'° Anin vitro wear test found that Empress wear resis-tance was comparable to that of enamel." Themarginal integrity has also been found to be clini-cally acceptable, thus confirming clinical and lab-oratory observations.̂ -^

The purpose of this investigation was to clini-cally evaluate Empress crowns.

Materials and Methods

Study Group

The study population included 55 patients (36women and 19 men, mean age 38.5 years) whowere treated using 144 crowns in the authors' pri-vate practices. Patients with severe parafunction,periodontitis, serious gingival inflammation, poororal hygiene, or high caries rates were excludedfrom this study.

Erom a minimum of 3 months and a maximumof 6 months, patients returned for oral hygiene re-call appointments depending on their periodontalcondition. As a sign of absence of inflammation,no bleeding on probing was required for inclusionin thÍ5 study. Old failed restorations with subgingi-val margins and gingival inflammation were alsoincluded in the study after being treated by crownlengthening procedures. Patients were informed ofthe purpose of the research and agreed to returnfor periodic recall examinations. Restorations in-cluded 51 central incisors, 38 lateral incisors, 12canines, 28 premolars, and 15 molars (Tables 1and 2|.

In 90 restorations (62.5%), the opposing denti-tion consisted of natural teeth, whereas 52 restora-tions (36.1%) were opposed by ceramic materials.Two crowns (1.4%) occluded with amalgamrestorations in the opposing arch. Seventy-one of144 examined teeth required a crown lengtheningprocedure because existing restorations presentedsubgingival margins, wbich violated the biologicwidth. A 6-month healing period was allowedprior to finalising the restorations. One hundredeight of the restorations were placed on endodonti-cally treated teeth, and the remaining 36 restora-tions were placed on vital teeth. Endodonticallytreated teeth received amalgam, resin composite,or a gold post and core.

The Inlernaiional lournal of Proillioclonlii 242 Í 10. Number 3.1997

Fr,ideani/A(|uila • With Empre

Tooth Preparation

For each restoration, the shade was determined priorto tooth preparation. A 90-degree rounded axiogingi-val line angle of 1.2 to 1.5 mm deep was circumfer-entially created to ensure maximum resistance formusing rotary diamond and hand instruments.'^

When greatly scalloped gingival architecture waspresent, a deep chamfer finishing line was used.'-iAll internal angles were rounded to reduce the stressconcentration during cementation and function. Aminimum thickness of 1.5 mm was maintained inthe occlusal surfaces of posterior teeth, and a mini-mum lingual thickness of 1.2 mm was maintained inthe maxillary anterior region. In this area, a well-defined lingual concavity was established to providea correct anatomic anterior disocclusion. The taperof the preparation was between approximately 5 lo10 degrees, depending on the abutment length.'^-""'

The location of the gingival margin was carefullyselected for each single restoration. In 33 out of 43posterior restorations the margin was located supra-gingivally. This facilitated impression making andevaluation of marginal adaptation while aiding inmaintaining periodontal health.

In the anterior area, margins were located at thegingival crest or slightly into the sulcus.''-"'

For equigingivally and intrasulcularly positionedmargins, gingival displacement was obtained using agingival displacement cord #00 (Ultrapack, Ultra-dent, Salt Lake City, UT) soaked with hemostatic so-lution (Hemodent, Premier Dental Products, Norris-town, PA). No displacement was needed insupragingivally prepared teeth.

After cord removal, the final impression wasmade using a polyether material (Permadyne, ESPE,Norristown, PA); the single impression-double mix-ing technique was followed using a light-activatedcustom-tray (Palatray LC, Kulzer, Wehrsbeim,Germany). An irreversible hydrocolloid (jeltrate.Caulk, Dentsply, Milford, DE) impression of oppos-ing dentition was made, interocclusal registrationswere recorded, and a facebow was used to relatethe master casts to a semiadjustable articulatorIDenar Mark II, Denar, Anaheim, CA),

The provisional restoration thickness was checkedto confirm the amount of tooth reduction. For the an-terior restorations, an impression of the provisionalrestoration was made to serve as a prototype for thefinal restoration using a silicone matrix technique. Allthe crowns were fabricated by the same dental tech-nician following the manufacturer's instructions andrecording occlusal and axial thickness. Ninety-threecrowns were made using the layering technique, and51 were made using the surface-coloration method.

The provisional restorations were luted using aeugenol-free material (Ereegenol, CC International,Scottsdale, AZ). Upon removal of the provisionalrestoration, the tooth was cleaned using a nonfluo-ridated cleaning paste (Syntac cleansing paste,Ivoclar, Amherst, NY).

The individual crowns were evaluated in termsof proximal contacts, occlusal relationships, shadematch, contour, and marginal adaptation. Minimalocclusal adjustments were performed prior to ce-mentation, leaving final occlusal verification untilafter cementation.

Luting

The restorations were etched with 4.5% hydrofluo-ric acid (Porcelain Etch, Ultradent) for 2 minutes,washed with water, and dried. A silane agent(Monobond S, Ivoclar) was then applied andblown dry. Concurrently, a dentinal adhesive(Syntac, Ivoclar) was applied to the prepared tooth.A dual-polymerizing resin composite cement wasused for luting the majority of the restorations.Sixty-nine (48%) were seated using Dual (Ivoclar)and 70 (49%) were luted using Variolink (Ivoclar).The excess cement was removed with a brush anddental floss interproximally. The margins of crownswere covered with glycerin gel, and the resin com-posite cement was light polymerized from eachside for 40 seconds. A gingival cord was placedinto the sulcus prior to luting to permit total re-moval of the residual cement from the crevice.Occlusion was again evaluated.

In five anterior restorations (37o), a zinc phos-phate cement (DeTrey Zinc, DeTrey/Dentsply,Weybridge, Surrey, England) was used to cover thedark abutment and adequately match the final color.

Evaluation

Photography and data forms were used as methodsof documentation in this study. Patients were re-examined by the authors at 6-month intervals forthe first year, and annually thereafter using a mirror,a sharp explorer, radiographs, and clinical slides;the absence of inflammation was ascertained usingthe no-bleeding-on-probing parameter at each ap-pointment. In failed restorations, the evaluators at-tempted to ascertain the cause of failure.

Color match, marginal discoloration, recurrentcaries, contour, and marginal integrity were evalu-ated with the modified US Public Health Servicecriteria'^ at baseline (placement) and at subsequentrecall appointments (Table 3). Kaplan-Meier^" sta-tistics were followed to calculate the survival rate.

243 The Internal I ona I lournai oí Prosthodontii

Clirical Experience With Em

Table 3 Criteria tor Clinical Evaluation ot Crowns

Color match Alpha (A) The restoration appears to match the shade and translucency of adjacent tooth tissues^Bravo (B) The restoration does not match the shade and translucency of adjacent toolh tissues, Dut ttie

mismatch is wilhin Ihe normal range of tootb shades,Charlie (C¡ The restoration does not match tbe shade and translucency of the adjacent tooth structure,

and tbe mismatcb is outside the normai range of tootb shades and translucency,Oscar (O) Tbe restoration cannot be examined witbout using a moutb mirror.

Marginal discoloration Alpba (A] Tbete is no visual evidence of marginai discoloration different from tbe color of tbe restorativematerial and from Ihe color of the adjacent tooth structure.

Bravo (B) Tbere is visual evidence of marginal discoioration at the junction of tbe tooth structure and therestoration, but the discoloration has not penetrated aiong tbe restoration in a pulpai direction.

Cbariie (C) Tbere is visual evidence of marginal discoloration at the junction of tbe tootb structure and therestoration that has penetrated along tbe restoration in a pulpal direction.

Recurrent caries Alpha (A) There is no visual evidence ot dark, deep discoioration adjacent to the restoration.Bravo (B) There is visual evidence of dark, deep discoloration adjacent to the restoration (but not directiy

associated with cavosurface margins].Contour (wear) Alpha (A) The restoration is a continuation of existing anatomic form or is siightiy flattened. It may be

overcontoured. Wben the side of the explorer is placed tangentiaily across the restoration, itdoes not touob two opposing cavosurfaoe iine angles at tbe same time.

Bravo (B) A surface concavity is evident. When tbe side of an explorer is placed tangentiaily across therestoration, the expiorer toucbes two opposing cavosurface iine angles at the same time, butthe dentin or base is not exposed,

Charlie (C) Tbere is a ioss of restorative substance so that a surface concavity is evident and tbe baseand/or dentin is exposed

Marginai integrity Alpba (A) The explorer does not catcb wben drawn across tbe surface of the restoration toward tbetootb, or, if the expiorer does catcb, there is no visible crevice along the periphery of tberestoration.

Bravo (B] The explorer catches and tbere is visible evidence of a crevice, into wbich tbe explorer pene-trates, indicating that the edge of tbe restoration does not adapt ciosely to tbe tootb structure.The dentin and/or the hase is not exposed, and tbe restoration is not mobile,

Charlie (C¡ The explorer penetrates a crevice defect that extends to the cementoenamel ¡unction.

Fig 3 Fracture occurred in a mandibuiar second molar.

Results

The 144 IPS Empress crown restorations were eval-uated from a minimum of 6 months to a maximumof 68 months with a mean period of 37 months.

One hundred one anterior and 43 posteriorrestorations were placed. The initial gingivalhealth, confirmed by the absence of bleeding onprobing, was maintained during all the scheduledrecall appointments.

Of 36 vital teeth, one tooth exhibited mild postop-erative sensilivity which disappeared after 1 month.

Five of the 144 crowns failed; two failures occurredbecause of a breakage in the resin composite core.The other three failures resulted from ceramic frac-ture. One occurred in a maxillary central incisor witha palatal thickness of 0,9 mm. The other two failuresoccurred on mandibular second molars where an oc-clusal cross section fracture was detected (Fig 3¡,These units bad a mean occlusal thickness of 1,5mm, and a minimum thickness of 1.2 mm in oneand 1,3 mm in the uther. Four of tbe five restorationswere replaced witb metal ceramic crowns. Only onecrown on a central incisor was replaced with anotherEmpress crown and included into the study where itis still present after almost 4 years.

Using tbe Kaplan Meier metbod, the estimated suc-cess rate, after almost 6 years, was 95,35% (Fig 4).

In comparing the anterior area (from canine tocanine) to tbe posterior area, no significant differ-ences were found between anterior and posteriorteeth according to the Kapian-Meier analysis (Fig5), as only two failures were recorded in the ante-rior segment as opposed to three in the posteriorone. All failures occurred in restorations luted usingDual, although this luting material did not performsignificantly differently than Variolink according tothe Kaplan-Meier analysis (Fig &). Three of the fivefailures occurred in crowns fabricated using tbe lay-ering technique, and two occurred in crowns fabri-cated using the surface-coloration method.

The Interrational lOJmal ol Proslhodorlii 244 Volume 10, Number 3, 1997

Fig 4 Kaplan-Meier survival-type curves fcr 144 tested crownsafter 5.5 years (95,35% confi-dence interval).

Frarteanl/Aqiiiiano

100

95 -

9 0 -

a> 8 5 -

O)

S

§ 80 -

°- 75-

70-

Cliiiicni Experience Witii Empress Crowns

Upper 95% confidence bound

I Ali crcwns (n = 144)

Lower 95% confidence bound

13

Time (y)

Fig 5 Kaplan-Meier suryjyal-type curyes for 144 anterior andpeste ricr positicned crowns.

rcen

tage

CL

100 -

95 -

90 -

85 -

80 -

75 -

70 -

C 1 2

^ ^ ^ Aniencr teeth (n =

• KHH Posterior teeth (n

3 4Time (y)

101)

= 431

5 6

Fig 6 Kaplan-Meier suryival-type cun/es for Dual luted crownsand Variclink luted crowns.

S»

Per

cent

9 5 -

9 0 -

8 5 -

8 0 -

7 5 -

70 -

1

Í

1

—

2

" Variclink cement (n = 70)

• Dual cement (n = 69)

3 4Time (y)

5 6

10 Number 3,1997 2 4 5 The Internjtionsl lournai of Prostliodontics

Clinical Expe rr.ldeari/AqLilano

Table 4 Results of the Clinical RatingsAccording to the Modified US Public Health ServiceCriteria

Table 5 Percentage of Dual and VarioNnit LutedCrowns Related to Marginal Discoloration According tothe Modified US Public Health Service Criteria

Category

ContourMarginal integrityMarginal discolorationColor rratchRecurrent caries

Alpha

97,895.092.894.2

100.0

Baling (%)

Bravo

2,25.07,25,80,0

Chariie

ÜU0.00,00,00,0

Type of cement

DualVariolink

Alpha

87,597,t

Rating (%)

Bravo

t2.52.9

Cfiarlie

0.00.0

Results concerning the clinical ratings of the re-maining restorations according to the modified USPublic Health Service criteria are shown in Table 4,Recurrent caries was not detected around any ofthe restorations. Contour quality was satisfactory,and proximal contacts were well maintained.

Two mandibular incisors showed insignificantincisai chippings as a result of accidenfal traumathe patient had experienced. The breakage did notinvolve the tooth-crown interface, and the restora-tions have survived for 2 years, A crack, visibleunder incident light, was detected in one centralincisor immediately affer cemenfation. With thepatient's informed agreement, it was decided tomaintain the restoration in the mouth. At an 18-month-recall the restoration was still In the samecondition and, as no signs of discoloration and mi-croleakage were detected, this restoration was notconsidered to have had an unfavorable outcome.

Discussion

The clinical results of this study on of IPS Empresscrowns were largely satisfactory. Five failures wererecorded during the observation period, three ofwhich resulted from ceramic breakage and twofrom core failure leading to abutment fracture at themarginal line without involving the ceramic restora-tion. Of the three breakages, the one that occurredin an anterior tooth was thought to be the result ofreduced thickness (0.9 mm). Thickness was notconsidered to be adequate (J ,2 mm) in the occlusalportion in the other two failed restorations.

In accordance with other clinicians,'^-^' the au-thors suggest that a 1,2-mm marginal shoulder fin-ishing line and an occlusal thickness of minimum1,5 mm are necessary fo ensure adequafe ceramic

thickness. What appears to be of interesf was thatfwo of the four restorations placed on second mo-lars failed.

Although the difference befween anterior andposterior restorations was not significant (P= ,14)asillustrated using the Kaplan-Meier "su rv i va I-type"curves (see Fig 5), it can be said that the esthetic ad-vantages derived from the placing all-ceramic ma-terials in posterior areas, where occlusal load ismuch more pronounced, does not justify the risk offracture. Therefore, in the absence of ideal clinicalconditions, the placement of a metal ceramicrestoration is recommended.

The survival rate illustrated with tbe Kaplan-Meier curves for Dual and Variolink luted crowns(see Fig 6) was not significantly different (P^ ,71),although five fractures occurred with Dual lutedrestorations. There appeared to be no relationshipbetween fracture rates for layered or surface-colored restorations.

On the basis of the criteria used (modified USPublic FHealth Service Criteria), a great percentage ofcrowns was Alpha for contour, marginal integrity,color match, and recurrent caries (Table 4¡; applyingthese criteria a clinical success of a restoration albeitwith some reservations could be established, al-though it was not possible to examine interproximalcrown margins using an explorer,--

Recurrent caries and contour were satisfactory,and color match was acceptable. Marginal in-tegrity was also satisfactory considering that mostof the margins were not visible and only 20,5% oftested crowns had supragingival margins. Marginaldiscoloration recorded the lowest Alpha rating(92,8%1. Marginal discoloration was Bravo in12,5% of Dual luted crowns compared to 2.9% inVariolink luted restorations (Table 5),

The International lourral of Prostliodontii 246

Frarfeiiri/Aquilano al Espe ri í With Empress Crc

Marginal discoloration was sometimes associ-ated with a decrease in marginal integrity and wasthought to be related to the cement used. The finalscore might have been influenced by the more re-cent introduction of Variolink (March 19921, al-though the total number of Variolink luted crowns(49%) was almost the same as those luted usingDual (48%)- Only 3% of the units were luted usingzinc phosphate cement. Stains on the accessibleareas of supragingival margins could usually be re-moved using a finishing bur.

The results of this study support the use of theIPS Empress pressed glass-ceramic restorations forsingle crown restorations, especially in the anteriorarea. This all-ceramic system has excellent translu-cency, improved strength, natural wear and resis-tance, and provides the clinician with an estheticalternative to the traditional metal ceramic crowns.

The postplacement period is still short, and con-tinued evaluations are necessary to fully ascertainthe results of this study.

Conclusions

One hundred forty-four IPS Empress crowns wereevaluated for a minimum of 6 months and a maxi-mum of 68 months with a mean period of 37.1months. The restorations were largely satisfactoryboth clinically, according to the modified USPublic Health Service criteria, and statistically, ac-cording to the Kaplan-Meier survival rate. Whileacknowledging that the restorations were placedand evaluated by the same individuals, the follow-ing conclusions may be made:

1. The Kaplan-Meier survival rate was 95.35%.2. The Kaplan-Meier survival curves did not show

significant differences between anterior and pos-terior teeth (P= .14). However, two crowns outofthe four units placed on second moiars failed.

3. Kaplan-Meier survival curves indicated that twoluting agents (Variolink and Dual) performedsimilarly (P = .71). However, all five fracturesrecorded occurred in crowns luted using Dualluting material.

4. Contour, marginal integrity, marginal discol-oration, color match, and recurrent caries wereall satisfactory in most observations according tothe modified US Public Health Service criteria.

References

1. McLe.in JW, leanîonne EÉ, Cliiciie GJ, Piiiauit A. Ail-ceramiccrowns and foi l crowns. In: Chiche G|, Pinaull A [eds].Esthetics of Anter ior Fixed Pros thodon l i c i . Ch icago:Quintessence, 1994:97-113.

2. Kappert HF, Knode H, Manzotti L. Metallfreie brücken furden sei tïen zahn be reich. Dem Labor 199C;2:177-183.

3. Kappert HF. Knode H. In-Ceram auf dem prufs land.Quintciîenz Zahnlech 1990;16:980-1002.

4. Paul S), Pictrolion N, Schärer P. The new In-Ceram SpinellSystem—A case report. Int | PeriodonI ResI Dem 1995;15:521-527.

.1 Adair P], Grossman DG. The castable ceramic crown. Int |Periodont Resl Dent 1984;2:33^5.

6. Maiament KA. The casi glass-ceramic crown. In: Preston |D(ed). Perspectives in Denial Ceramics. Chicago: Quintessenta,1983:331-342.

7. Hankinson JA, Cappetta EG. Five years' clmicai experiencewi lh a leucite-reinforced porcelain crown system. Int JPeriodonI Resl Denl 1994;14:I39-1S3.

8. Fradeani M. Barducci G. Versatility of Empress reilorations.Part I: Crowns. | Esthet Dent 1996;3:3O-3B.

9. Dong |K, Lulhy H, Wohlwend A, Schärer P. H eat-pressed ce-ramics: Technology and strengti i . Inl | Proslhodont1992;5:9-16.

10. Lehner CR, Schärer P. All-ceiamic crowns. Curr Opin Dent1992;212¡:45-52.

11. Krejci I, Lut! F, Reimer M, Heinzmann |L. Wear of ceramicinlays, iheir enamei antagonists, and lu l ing cements. |Prosthet Dent 1993;G9:425^3O.

12. Heinzmann |L, Krejci I, Lutz F. Wear and marginai adaptationof gla55-ceraiiiic inlays, amalgam and enamel [abstraci 423]. ]Dent Res 199O;69lspecia! ¡ssue):161.

13. Friedlander L, Munoz CA, Goodacre C|, Doyie MC, MooreBK. The effect of tooth preparation design on the breakingstrength of Dicor c rowns. Part 1 . Int | Proslhodont199O;3:159-168.

14. Maiament KA, Grosmann DC. The cast-glass ceramic reslora-lion.J ProslhelDenl1987;57i674-6B3.

15. Doyle MC, Coodacre CJ, Munoz CA, Andres C|. The effect oflootb preparation design on the breaking strength of Dicorcrowns-Part2. IntlProsthodont 1 9 9 0 ; 3 T 2 4 1 - 2 4 8 .

16. Doyle MG, Goodacre CJ, Munoz CA, Andres G|. The effect oftootb preparation design on the breaking strengtb of Dicorcrowns. Part 3. Int j Prosthodont 199O;3:327.-34D.

1 7. Caigiulo AW. Dimension and relation of the denlo-gingivaljunction in humans. J Pefiocfonlol 1961 ; i2:2h]-267.

18. Ricbter WA, Ueno H. Kelalionship of crown margin place-men! Lo gmgival inflammation. | Prosthet Dent 1973;3O:156.

19 Ryge G, Cvar |F. Criteria foi tbe Clinicai Evaluation of DentalRestorative Matenals. US Public Health Service publication No.790-244. San Francisco: Government Printing Office, 1971.

20. Kaplan EL, Meier P. Non parametric estimation from incom-plete observations. J Am Statist Assoc 1958;53:4S7^65.

21. Grossman DG. Casi glass ceramics ceramic. Denl Clin NorthAm 1985;29:725-737.

22. Cbrislensen C|. Marginai fil of gold inlay castings. ] ProslbetDent 1966;! 6:297-305.

•10, Number 3,1997 247 The Internationöl loutnal oí Prosthodontil