Precision Attachment-RetainedRemovable Partial Dentures:

Part 2. Long-Term Studyof Ball Attachments

Bengt Öwall, Odont DrProfessor and ChairmanDepartment of Prosthetic DentistrySchool of DentistryFaculty of Health SciencesUniversity of CopenhagenCopenhagen, Denmark

Patients provided with 24 ball attachment-retained removable partialdentures were followed up to 23,5 years. The prosthetic treatment includedfixed partial dentures in 8 arches, and combinations of crowns and splintbars in 16 arches. All ball attachment matrices were supplied witb verticalocclusal stops in contact with the patrices, A nonresilient hinged couplingwas established between fixed partial dentures and removable partialdentures. Seventeen arches had only two or three remaining teeth, A total of65 abutments for fixed partial dentures were included, 30 of which were rootcanal treated and supplied with posts. No technical failures (loss ofretention/cement failure, root or tooth fracture, metal framework fracture)witb tbe fixed partial dentures were recorded, Witb tbe removable partialdentures, additional retention with clasps was introduced in 2 of the dentures(in addition to 8 original ly], 2 had to be remade after fractures, and 4dentures were relined. Int j Prosthodont 1995:8:21-28.

C ombinations of fixed and removable partialdentures are used in many different ways.

Crowns, and fixed partial dentures (FPDs) whichrestore parts of the remaining natural dentition orare used as abutments for clasp-retained remov-able partial dentures (RPDsl are usually notincluded in the trealment alternatives that areknown as fixed/removable partial dentures. Theterm generally refers to the use of attachments,double crowns, bars, etc. Overdentures with rootattachments are sometimes included in the termsattachment-retained or precision attachment-retained dentures,'"' The double conical crown isalso a technical alternative combining fixed andremovable prostheses,'"" In some of the rather fewstudies that have been published regarding preci-sion attachment-retained dentures, different typesof attachments have been grouped in the samereport."

Reprint requests: Or BengI Ownll Department of ProithelicDentistry, School of Dentiilty, Faculty of t-iealth Sciences,University of Copenhagen, 20 Norre At!é, DK-2200Copenhagen N, Denmark.

A distinction between rigid and resilient combi-nations is made by Rantanen et al,= who recom-mended tbat resilient connections should beavoided, as they sbowed a higher failure rate thanrigid designs. Any clinical long-term studies ofhinged or stress-distributing combinations are notavailable,

Tbe conical crown/double crown prostheseshave a rigid connection between the fixed andremovable segments, and good results have beendemonstrated.'"" ÖwalT' studied FPD/RPD combi-nations having only a rigid slide attachment; hereported good results with tbe RPD but failures,including tooth fractures, cement failures, andframework fractures, with the FPDs, That studytuuld not determine if the rigid connection of thedistal-extension RPDs increased tbe risk of techni-cal failure. All types of fixed prosthodontics forcompromised dentitions are at risk of teebnicalfailure,'""'- When comparing the results of thosestudies, tbe addition of an RPD has apparently notincreased the risk of failure. There is, therefore,reason to study prosthetic designs using hingedconnections of the removable to fixed segment.

VolumeB, Number 1, 1995 The International lojmal of Prosthodontics

Precision Attiichment-Relaineil Rtmovable Parlial Dentures: Pari 2

Table 1 Dental and Denture Status of the ArchesOpposing Those Treated with Ball Attachment-retainedRPDs(n = 24)

Treated a roh

Opposing Arcb

Natural dentition cnlyRPDComplete dentureBall attachment RPD

Table 2 Types of Fixed Components, and Numberand Endodontic Status of Abutments

Restoration Number TypeNumber

abutments

Numberabutmentsendodontic

post

{1 ball attachment,conventional clasps oontralateral)

3 Two-unit splint 6

¡2 ball attachments)5 Splint or FPD 27

16 Bar 39

especially ¡n regard to prognosis, failure rate, andpatient satisfaction.

The purpose of this investigation was to examinea group of patients treated with one type of nonre-silient, hinged attachments—ball attachments—forthe connection between FPD and RPD fabricated byone prosthodontist.

Materials and Methods

in part related io their denture problems- '^j^^loss and the position of the remaining teet ,̂was the indication for the initial t̂ eatmen^^^^was the indication for thepermit treatment with a fixed Fand the amount oí remainindirect restoration and ''6'^^' , ,^gptto permit service as an RPD a '^

precludedrestorations

Arches Treated

The number P.ace.ent

1:^stre^ngthof the

load from a rigid slide P ^ « ^ ' . ^ ' ^ fTherefore, a hinged connection betweenand the removable prostbeses was cbosen to pro-vide some stress distribution.

The reasons for tbe tooth loss m the archestreated using the ball attachment constructionswere: trauma (3 arches, 2 maxillary, 1 mandibu-lar, patient ages 76, 3 1 , 24 years); caries (5arches; patient ages 57 to 74 years, mean 67.6);periodontal disease (6 arcbes; patient ages 43 to66 years, mean 58,5]; and a combination ofcaries and periodontal disease [6 arches; patientages 60 to 84 years, mean 72,7|, Caries as used inthis context refers to large lesions, teeth withmany restorations, root f i l l ings, dowel postsand/or tooth or root fractures fol lowing veryextensive conservative treatment, and periapicalpathosis as a result of caries or previous restora-tion. In 4 of ihe arches (patient ages 64 to 77years, mean 71.0), the original reason for thetootb loss could not be established.

The material consisted of 24 FPDs with ballattacbments in 21 patients. Three patients had suchconstructions in the maxillae, 15 in the mandible,and 3 in both arches. Twenty-one of the restora-tions included 2 ball attachments, and 3 included1 attachment each. Tbus, 6 maxillary and 18mandibular prostheses were included. Tbesepatients represented all of tbose treated by theauthor over a period of 30 years (1964 to 1993)using non resilient ball attachment constructions.

Patients

For the 21 patients (10 women, 1 1 men) aged 24to 84 years (median 66 years) at ihe time of the ini-tial treatment each arch was considered individu-ally (n = 24),

Two patients, one with maxillary and mandibu-lar restorations, and one with a mandibular restora-tion only, were undergoing psychiatric treatment.

Opposing Arcbes

The arcb opposing the arch receiving the pros-thesis was always treated completely during thesame period. All patients except one maintainedtbe same status in the opposite arch over the studyperiod. This one patient was fitted with a ballattachment-retained RPD in the maxillae opposinga naturally dentate mandible. After 10 years, a ballattachment-retained mandibular RPD was placedfollowing extractions necessitated by periodontaldisease. This was recorded as two ball attach-ment-retained RPDs. The status of the opposingarches is presented in Table 1,

Abutmertt Teetb and FPD

The distribution of abutments In the archestreated and the types of fixed components are pi-g_sented in Table 2,

The Inte'iiational lournal of Prosttiodonlii

I Atta cil mem-U eu i ned Me able Partial Dentures: Part 2

Fig 1 Ball attachment matrix with metai vertical stop at the9.5 year recall. The denture was supplied with additionalclasping alter 9 years.

Fig 2 Cast bar spiinting abutments with bali attachment onthe distal surfaces at 9.5-year cheok-up.

Three arches had two splinted crowns unilater-ally; two of these arches were treated followingtrauma and the other following loss of teeth fromperiodontal disease. One of the prostheses was aunilateral maxillary obturator following a shotgunaccident, and the bail attachment was suppliedwith a locking device.

Of the other restorations, five were conventionalfixed partial dentures with pontics or splints (therewere 6 unit FPDs, one was a 5-unit FPD, and onewas a 4-unit FPD), and 16 incorporated an anteriorbar connecting and splinting the abutment teeth.The bars that spanned an edentulous space of 4incisors in 12 restorations (3 maxillary and 9mandibular) and 3 incisors in 4 prostheses (1 max-illary and 3 mandibular) were also used for indirectretention of the RPDs.

Edentulous Areas

Twenty of the RPDs were bilateral distal exten-sion ¡3 maxillary and 17 mandibular), 2 were max-illary unilateral distal extensions, one was an obtu-rator, and 2 were large, anterior replacements.

Marginal Bone Support

The marginal bone support of the abutment teethof the fixed partial dentures, as indicated on intra-oral radiographs when treatment was initiated, was80% to 90% in 4 arches; 60% to 70% in 9 arches;40% to 50% in 10 arches; and .'30% in 1 arch. Thetooth surface with tbe most reduced marginal bonesupport was used to characterize the arch, as mostarches included very few abutment teeth and allwere essential for the survival of the FPDs.

Bali Attachments

Gold Roach-type solid ball attachments measur-ing 2.5 mm in diameter with patrices (Sjodings Aß,Solna, Sweden) or 2.5 rtim diameter patrices withan activation slot in the ball (Cendres & MétauxSA, Biel, Switzerland) were used. The respectivelengths of the matrices were 4.0 and 4.Ö rtirn.

In all the restorations, the matrices were suppliedwith a gold vertical occlusal stop and were nonre-silient, ie, the balls were in contact with the verticalstops when the relationship between the fixed andremovable partial denture was established (Fig 1 ).

Prosthodontic Technique

The fixed prostheses were made according togenerally accepted principles for retention, occlu-sion, and vertical dimension. The bars were madeeither of prefabricated, round metal bars (n = 4),curved to fit the residual ridge and provide suffi-cient gingival relief at the proximal surface of theabutment teeth and soldered to the crowns, or ofindividually shaped (n = 12), cast metal connectors{Fig 2).

All removable partial dentures had cobalt-chromium frameworks (Vi tal i ium, Austenal,Chicago, IL|. In the maxillae, al! major connectorswere metal palatal plates. In the mandible, 16were cingulum bar connectors resting on the FPDand not covering gingival tissue, and 2 were lin-gual bars, owing to space problems, as the FPDshad short clinical crowns.

Distal extension bases were extended to includeeither the retromolar pad or the tuberosity. All arti-ficial teeth in the RPDs were of acrylic resin.

The patrix attachment of the balls was mounted

Volumes,Numberl,1995 23 The irlernalional lojrnal of Hrosthodoniics

ifon AltElclimenl-Ri't.iineil Removable l'ártial DcnLures: pjri 2

Fig 3a Ball patrices were placed in tbe incisal parts of thecrowns to aliow matnces to remain long and to provide gingi-val reliet (9 5-yeaf recall).

Fig 3b Same patient with denture in place.

in the incisal/coronal part of [he crowns to pre-serve the matrices' length and to provide adequategingival relief on (he attachmenf side of the crowns(Figs 3a and 3bl-

All three of the 2-crown splints were suppliedwifh auxiliary clasping (for the obturator patient inthe form of a locking device! at the time of initialplacement. Of the 5 conventional FPDs, 2 were ini-tially supplied with clasps, and of the 16 FPD barconstructions, 3 were initially supplied with claspsand one with Ipsnclips (Cendres & Métaux SA) ontothe bar. No other retention to the bars was used.

Recall and Observation Period

The patients were followed by individual recall,usually at 6-month intervals. Some patients did notrespond to the recalls and were then recorded at thetime they returned. Three patients consulted otherdentists in the area where they lived, from whomprofessional information could be obtained for thisstudy. Patients who died, and one patient whomoved from the district and with whom contact waslost, were recorded for that part of the observationperiod in which they could be evaluated.

The observation period was calculated inmonths from the day of insertion of the prosthesisup to the las! clinical evaluation. For patients wholost their restorations as a result of extraction,obtained alternative trealment, or abandoned theirRPDs, the observation period terminated. All otherpatients were followed continuously and compli-cations, if any, were recorded.

Results

Observation Period

The observation period ranged from 14 to 282months (median 89). For 9 prostheses, the observa-

tion period was terminated because the patientdied or moved from the district. One patient wasfollowed for 90 months, after which contact waslost. Four prostheses were lost, 3 as a result ofextraction of the abutment teeth necessitated byperiodontal disease, and 1 because the patientabandoned her mandibular RPD and did not get itadjusted for re-use (see section below on patientsatisfaction).

Two RPDs in one patient had to be abandonedand the bar constructions altered for fixed partialdentures after ñ months (see section below onpatient satisfaction]. Eight prostheses, includingtwo that were observed by other dentists, are stillunder observation (see Table 2). The distribution ofthe treatments over tbe 30-year period is: first 10years—10, second 10 years—8, and last 10years—6.

Patient Satisfaction

All removable prostheses included anterior teeth,obligating their use for social/esthetic reasons. Ninepatients had complete dentures in the opposingarch; the others had their natural dentition, with noremovable denture (5 patients) or with a removablepartial denture (7 patients). All but one, however,found the prosthesis satisfactory and well function-ing. This patient had also had severe denture prob-lems before and was under psychiatric care at thetime of prosthodontic treatment. Maxillary andmandibular ball attachment-retained RPDs weretried for 8 months. However, the patient could nottolerate the removable dentures and, in spite of avery reduced number of abutments, superstructureswere fabricated to modify the bar constructions tnfixed partial dentures. The observation period fthis patient was thereupon terminated. °'^

One other patient was also undergoing psvrh'atric care for denture problems at the time of nr

The Internationai loiProsthodonti. 24 Volumes, Number 1, 1995

Precision AUathmcnt-ííctaincd Re able P^rli.ii Dentures: Part 2

thetic treatment. He had a complete maxillary den-ture and an overdenture in the mandible retained byroot attachments on endodontically treated canines.He was provided with crowns on the canines and aconnecting bar, two ball attachments and an RPD.This was used for 117 months until the canines werelost as a result of periodontal disease.

One patient, 81 years old at ihe time of initialtreatment, was provided with crowns on twomandibular canines and one lateral incisor, a con-necting bar, two ball attachments, and an RPD.She used this RPD satisfactorily for 42 months,over which time she became severely senile andstopped using her RPD. The observation periodwas then terminated.

None of the patients experienced any problemsin manipulating their RPDs when placing orremoving them.

Retention and Stability

All patients were asked about retention at thefollow-up visits. Two patients received auxiliaryclasping after 69 and 164 months, respectively, asthey considered the retention unsatisfactory.Others were satisfied.

The slots in the balls were never used for activa-tion. This was avoided to prevent fracture, whichcould not be repaired since the balls were solderedto the FPDs.

The RPDs without any auxiliary clasping hadalmost no active retention in the path of RPD place-ment/removal, but they had excellent stabilityagainst horizontal displacemenl. This might accountfor the patients' adaptation and satisfaction.

Technical Maintenance

Relines. Four mandibular RPDs were relinedbecause of tilting around the hingesA)all attachmentsafter 40, 41, 74, and 90 months. At the same time,the artificial teeth were repositioned or occlusallycorrected. Re-establishment of the RPD-mucosarelationship was also undertaken at remakes.

Restorative Procedures. Seven crown marginsrequired restoration as a result of caries. One blindpatient received five restorations over the last 6years of her î8-year observation period, butretained her prostheses until her death at the age of81. Sixty-six abutment teeth in all were included inthe FPDs examined, and tbe mean observationperiod of 7.5 years indicates that caries control badbeen achieved.

In three RPDs, the acrylic artificial tooth housingtbe matrix fractured, and in one RPD all the acrylic

teeth had to be replaced as as result of wear (after99 months).

Remake and Major Corrections of RPD. Afterrepairs of two mandibular cobalt chromium frame-works had failed, they were remade to fit the origi-nal fixed partial dentures.

One patient was provided with a duplicate max-illary denture for social reasons; it was usedinstead of the intact original prosthesis.

In two mandibular restorations, the caninesbecame completely periodontally compromisedafter 30 and 42 months, respectively. The amountof marginal bone support on these two teeth wasvery reduced at the time treatment began, as aresult of gingival stripping caused by the previousRPDs. One of these patients rejected treatment. Theother had the bar shortened to a mesial arm extend-ing from the remaining tooth, and the compromisedtooth was extracted. The RPD was adjusted. After afurther 11 months, the retained abutment tooth alsorequired extraction, a complete denture wasplaced, and the observation period terminated.

No technical failures, including loss of reten-tion/cement failure, root or tooth fractures, ormetal framework fracture, occurred with the FPDs.

Prognosis

The patient who was unable to accept her 2RPDs had no technical failures during tbe 8months that they were in use. She was accordinglyexcluded from the material in presenting the tech-nical prognosis. Of the remaining 22 restorations,3 failed because of periodontal complications(after 4J, 54, and 117 months]- One of these hadhad one oí the two abutments extracted earlier andthe denture was altered, and one had had aremake of a fractured RPD. One further RPDframework fracture occurred and the denture wasremade. The prognosis is summarized in Fig 4 andsome examples are shown in Figs 5 to 7.

The patient who was offered a shortening of theFPD and an adjustment of the RPD is reported as apartial failure.

Minor repairs, relines, replacement of loosenedor worn resin teeth, restorations, and auxiliaryclasping did not affect the prognosis to any notableextent; consequently, they are reported under tech-nical maintenance.

Discussion

There are several indications for the use ofattachments as a coupling between an FPD and anRPD. Esthetics, meaning the avoidance of clasps, is

B, Number!, 1995 25 a I oí Pro&thodontu

'recision Atta chilien 1-Re íParl i i i l Dqnlurcí; P,irl 2

•erce

00-

90-

SO'

70

60'

50-40-

30

20-

0

n.

n=2l

ÈÊÊÊL-

fer

(t

n=20

anH

\

n=18

pVVwtm

-' \ I''- \

11=17

^ ^

• r3 years 5 years 10 ye

Observalion period

t 5 y e

Restoralion removed/denlurc abandoned

Movcd/üicd/lost

Fig 4 Long-term prognosis of fixed restorations and nonre-silient ball attachment-retained RPDs. Restoration removedindicates tiiat the FPD and all abutments were extracted.Denture abandoned indicates that patient did not tolerate theRPD. Lost indioates that oontact was lost with the patient, nindicates the nurnber of patients that could have reached 3-,5-, 10-, or 15-years observation as a result ot the time otplacement.

Fig 5 Recall radiographs (10 years 7 months) of mandibularFPD, including crowns on the canines, a splint bar, and ballattachments distaily. Same patient as in Fig 2 and 3. A maxil-lary complete arch FPD opposed the prosthesis.

often mentioned," In this study, tbat aspect of theiruse has not been considered important; rather,support (resistance to movement of the prosthesistoward the tissues), stabilization (resistance to hori-zontal movement of the prosthesis), and fixation(resistance to separation of the RPD and FPD) havebeen emphasized.

This study reports tbe results of one type of P' .sion altachment-retained RPDs used on a '•^'^'-,'group of patients treated by one prosthodonlisi- ^results are thus far from applicable to other gioupsof patients treated by other dentists. As th' i ' ' ^''^very few existing reports about precision aiucn-ment-retained RPD, and especially ball attach-ments, there is clearly a need to collect experiencesand results. One advantage of this material is thatthe same author has reported a material of distalextension removable partial dentures retained byrigid intracoronal slide attachments from the sameperiod of time.'̂ Comparisons of the two attachmentmodels are thus possible without the disadvantageof different clinicians being involved.

It must be noted that {!) the patient materialincludes a majority of elderly patients with compro-mised dentitions, (2) the marginal bone support ofseveral of the teeth included in the FPDs wasreduced, {3) many abutment teeth were endodonti-cally Irealed and posts were placed, and (4) twopatients were problem patients to such an extent thatthey were under psychiatric care. These facts must beconsidered when comparisons are made. Althoughabout 50% of tbe patients had a history of caries orcavilies at the time of prosthetic treatment, caries waswell under control. Only three patients developedcarious lesions in connection with the FPDs, two ofwhom had only one carious lesion each.

The overall technical results of the treatmentwith ball attachment-retained RPDs seem verygood. At 15 years the success rate was 60%, and at10 years it was 70%—if patients who died ormoved are distributed in failed and satisfactorygroups in relation to frequencies. This is muchhigher than tbe results presented by Vermeulen,'who found that the 50% survival time was 8 yearsfor a material of 183 precision attachment-retainedRPDs, including 40% ball attachments.

The material of rigid slide attachments presentedby Owall" had an extrapolated 50% survival timeof about 20 years, and a survival rate of about 65%up to 1 5 years. These figures are almost exactly thesame as those for ihis material.

When compared to the material of naid slideattachments reported by Owall,^ (here also seemsto be a similar freqt^ency of technical failures Theneed for relmmgs differed: 9,5% (two oostextr;,r-tion relines excluded] for the rigid sMd^and 16,6% for the ball attachment material

It is noteworthy, however, that in thisthere were 18 mandibular (75%] and 6restorations. In the rigid slide attachmentthe corresponding figures were 14(26%] and 39 maxillary prostheses,'

26 Volumes, Number 1, Î995

enl-Relained Removíhlc Portiol Dentures: Psrt 2

ig 6 Pretreatment (a) and tollow-up (b) radiographs 70 months later. The patient bad bis FPD and RPD un'lonths after treatment, A complete maxillary denlure was opposing.

Fig 6 .il his death, 96

A comparison of mandibular prostheses onlyreveals that, even though those retained with ballattachments had fewer abutments and a muchhigher frequency of root canal-treated abutmentswith root posts, there were fewer failures. It mustbe emphasized, however, that this comparison wasmade only for the mandible, where the FPDs inalmost all the patients were in normal incisor rela-tionship, ie, there was a horizontal overlap coun-teracting the risk of a frontal tilting of the FPD,

It is also noteworthy that the Roach ball attach-ment was the first type of attachment to be fabri-cated, and that it is still being produced andused."""

The results of Hultén et al,' who studied a con-ical crown/double crown-retained RPD material,indicate a high risk of failure. The failure rate intheir material was considerably higher than inthe present material when using ball attach-ments, even though the quality and distributionof tbe abutments for the crowns were similar inbotb studies. Conical inner crowns are separateand the splinting is obtained by the RPD. Thismethod does not give the same reliable and sta-ble splinting as that provided by bars used inconjunction with the ball attachments as bar-splinted FPDs with nonresilient extracoronal ballattachments.

Thus, it is essential to point out that the mate-rial presented herein includes only nonresilientextracoronal ball attachments on splinted abut-ments.

Fig 7 Patient with the technically weakest abutment teetb:two mandibular canines with root anchors on posts 1o supportan overdenture. The patient oould not tolerate the overden-lure. The construction above functioned tor 1t7 mcntbs, wbenIt was lest because ot periodontal complications and toothmobility, A complete maxiliary denture was opposing.

Conclusion

In this material of 24 RPDs coupled to FPDswith nonresilient Roach ball attachments, fol-lowed for periods ranging from 7 to 282 months,few technical failures were recorded. Two RPDframeworks were remade after fractures, 4required relining, and 2 (in addition to an initial 8)required auxiliary clasping. JNO technical failuresoccurred with the FPDs,

, Number i , 1995 27 The International Journal oi ProBthodontics

on Atliichment-Retamed Re able Panial Dentures: Part 2

References

1. Preiskel HW. Precision Aitachmenls in Dentistry. London:M Krimplon, 1979.

2. Ranlsnen T, Siirilä H, Yli-Uipo A, Màkilâ E, Investigationsof the therapeutic success with dentures retained by preci-sion altachrrents. il. Partial dantures. Proc Finn Dent SocI972;68:73-B5.

3. Mensor MC. Removíible partial overdentures wi thmeclianical IprecisionI attachments. Dem Clin Norlh Am1990;34:669-66l,

4. Heners M, Walther W. Klinische iJewarung derKonuskrone ais peiioprolhetisches Konstruktionseiement.Eine Langzeitsstudie. Dtsch Zahnartzl Z 1988;43]52S-529.

5. Ericson A, Nilsson B, Bergman B. Clinical results inpatients provided with conical crcwn retained dentures.Inl l Prosthodont 1 99O;3:5I3-521.

6. Hullén J, Tillström B, Nilner K. Long lerm clinical evalua-tion of conical crown lelained devices, Swed Dent |1993:17:225-234,

7. Vermeulen A Een Decennmm Evalualie van PárteleProthesen. Een Beschrijvcnd Klinisch LongitudinalOnderzoek |PhD thesisl. The Netherlands: CatholicUniversity of Nijmegen: 1 984.

Öwall

13

ß. Molin M, Bergman B, Hricson Â. A clinical ^

conical crown retained denlureí. J P ™ " "

l993;70:251-256. . g^iovafale parHal

3. Öwall S. Precision alt.ichmenl-relainedJ ^^^^^ ^^^ ^

dentures: Pari I . Technical long-te

Prcsthodont 1991:4:249-257. , combined

,0. Nvman S, I'-^SV/os'-hLr:."-"-/patients with

163-169. of service in fined

iqng 13185-189.öwall B, Almfeld, I, Heibo M. Twenty-year experiencew<th 12-unit fixed parliai dentures supported by two abut-ments Int ] Prosthodont 1 991 ;4:24-29.Henderson 0, Steffel VL. McCracken's Removable PartialProsthodontics, ed 6. St Louis: Mosby, 1981.Prosthoaonttcs, ffu • . J I LU^I.-. ; i

14 Goslee HI. Removable bridgework. Patented and manu-factured attachments. Denl Cosmos 1 SÜ6;28:644-665.

15. Roach FE. Conserving the natural teeth in supplying par-tial dentures. Denl Cosmos 19O8;3O:1 7-23.

16. Zinner IA. Precision attachments. Dent Ciin North Am1987;31:395-416.

of Proslhodonlic 28 Volumes, Numfaerl, 1995