Four Post-and-Core

Combinations as Abutments
Stefan Ellner, LDSa
for Fixed Single Crowns: Tom Bergendal, LDS, PhDb
A Prospective up to 10-Year Study Bo Bergman, LDS, PhDc

Purpose: In the present prospective study, four different post-and-core systems were
evaluated over a period of up to 10 years. Materials and Methods: Fifty endodontically
NO PART OF THIS ARTICLE MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM

treated teeth in 31 patients were randomized to one of four groups for post-and-core
placement: Group 1 received conventional tapered cast posts and cores (n = 14); group 2
PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.

received ParaPost system prefabricated gold posts with cast cores (n = 13); group 3 received
ParaPost system cast posts and cores (n = 13); and group 4 received Radix-Anchor posts (n
COPYRIGHT © 2003 BY QUINTESSENCE PUBLISHING CO, INC.

= 10). Clinical and radiologic evaluations were made. Results: One post and core in group
WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

2 was functioning well 58 months after placement when the patient died. No posts and
cores in groups 1 or 2 had been lost or had any complications, one in group 3 had been lost
because of a root fracture after 108 months, and two in group 4 had been lost after 54 and
88 months, respectively, because of loss of retention. The final treatment result for 46 of the
49 remaining posts (30 patients) was successful. The overall failure rate was 6%. There were
no statistically significant differences between the four groups. Conclusion: If recommended
procedures are strictly followed, posts and cores can serve as abutments for fixed single
crowns with satisfactory long-term results. Int J Prosthodont 2003;16:249–254.

A number of prefabricated and individually de-

signed posts have been developed and, in gen-
eral, marketed with no presentation of any evidence
light-curing resins, and one used cast-metal posts
and cores. Of 154 restorations, 11% failed. The au-
thors found no statistically significant differences be-
of their superiority or inferiority in relation to other tween the three techniques and concluded that ex-
systems. To evaluate the different posts, numerous in tended periods of clinical investigation are needed to
vitro studies have been performed.1 However, in determine which systems will have the best clinical
vitro studies often produce conflicting results and results in the long term.
thus are of little help in deciding which post-and-core The present prospective clinical study was de-
system would be best in a given clinical situation. signed to evaluate the clinical outcome of four dif-
In contrast with the numerous in vitro studies, ferent post-and-core combinations in a longitudinal
comparatively few clinical studies have been pub- perspective. The hypothesis was that the four com-
lished. Most are retrospective,2–17 but to the best of binations are comparable, provided that a 2-mm fer-
the authors’ knowledge, only one prospective study rule is prepared on the tooth.
has been published.18 That 3-year study compared
three techniques; two used prefabricated screw posts Materials and Methods
or ParaPost stainless steel posts (Whaledent) with
Fifty endodontically treated teeth in 31 patients (14
aSenior Consultant and Head, Clinic for Prosthodontics, Specialist women and 17 men aged 16 to 75 years, mean age
Dental Care Center, Kalmar County, Sweden. 48 years) met the inclusion criteria. All patients gave
b Head, Department of Prosthetic Dentistry, Institute for Post-
their written informed consent to participate in the
graduate Dental Education, Jönköping, Sweden.
cProfessor Emeritus and Former Chair, Department of Prosthetic study. The patients were randomized to a certain
Dentistry, Medical-Odontological Faculty, Umeå University, Sweden. type of post-and-core combination by lot. Initial
prosthodontic treatment was performed between
Reprint requests: Dr Stefan Ellner, PO Box 299, SE-391 23 Kalmar,
Sweden. Fax: + 46 480 84977. e-mail: stefane@ltkalmar.se
September 1989 and September 1993.

Volume 16, Number 3, 2003 249 The International Journal of Prosthodontics

 Posts and Cores as Single-Crown Abutments Ellner et al

7 Group 4

No. of posts and cores

6 Group 3
5 Group 2
4
Group 1
3
2
1
0
15 14 13 12 11 21 22 23 24 25 45 44 43 33 34 35
Tooth No.

Fig 1 Distribution of posts and cores according to group (see Materials and Methods) and tooth
(Fédération Dentaire Internationale tooth-numbering system).
NO PART OF THIS ARTICLE MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM
PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.

Study Criteria Endodontic Treatment

Clinical inclusion criteria were: If the root filling of the tooth to undergo post-and-core
COPYRIGHT © 2003 BY QUINTESSENCE PUBLISHING CO, INC.
WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

buildup had a radiologically detectable poor seal, was

• One or more single-rooted teeth in need of a sin- less than 3 mm from the radiologic apex, or had an api-
gle crown and not intended for use in the support cal radiolucency near the tooth, the tooth was en-
of a removable partial denture (Fig 1). dodontically retreated. All root canals had been filled
• No marked signs of bruxism, expressed as heavy with gutta percha. The gutta percha was removed with
attrition. A limited number of wear facets on the a Gates reamer (Hager & Meisinger). The periapical
remaining teeth were accepted. conditions were classified according to Petersson et al19:
• At least 2 mm of remaining vertical marginal tooth
substance available for ferrule preparation. • Normal = Apical periodontal ligament space not
more than double the width of the other parts of
Some of the teeth had a screw post retaining an the root; appearance of the surrounding osseous
amalgam crown; none of these posts had been structure is normal.
placed more than 5 years before the tooth was in- • Bone destruction = Periapical radiolucency ob-
cluded in the study, and the screw post was re- served; bone destruction ≤ 5 mm and > 5 mm
moved. were registered separately.
At the initial examination, the following general • Periapical conditions not evaluable = Poor radi-
clinical and radiologic factors were evaluated: ographic quality; periapical structures impossible
to evaluate.
• Clinical and radiologic caries for all teeth.
• Sagittal, transverse, and vertical maxilloman- Post-and-Core Designs
dibular relationships.
• Occlusion, including intercuspal contact on teeth • Group 1 = Conventional tapered posts and cores
with posts and cores (stable/unstable). Any oc- cast in a type III gold alloy according to Zarb et
clusal interference was noted and adjusted. al20 (n = 14)
• Occlusion in lateral excursions (balanced/unbal- • Group 2 = ParaPost system with prefabricated
anced). Canine-protected articulation was regis- posts in nonoxidizing gold alloy and cast cores in
tered separately. a type III gold alloy (n = 13)
• Plaque at four surfaces: mesial, buccal, distal, • Group 3 = ParaPost system with cast posts and
and lingual, after use of a dye solution (yes/no). cores in a type III gold alloy after the use of pre-
• Pocket depth at six surfaces: mesiobuccal, buccal, fabricated burnout posts (n = 13)
distobuccal, distolingual, lingual, and mesiolin- • Group 4 = Radix-Anchor (Maillefer) threaded ti-
gual. tanium posts with cores in chemically cured resin
• Bleeding on probing of pocket depths (yes/no). composite (n = 10)
• Gingival retractions, gingival swelling, cracks,
and so on, around the experimental tooth. The type III gold alloys used were JS C-guld (JS Sjöding)
• Periapical conditions. Intraoral radiographs were and Protor (Cendres & Métaux). Concise (3M) and
made using the long-cone parallel technique. Adaptic (Dentsply) were the resin composites used.

The International Journal of Prosthodontics 250 Volume 16, Number 3, 2003

 Ellner et al Posts and Cores as Single-Crown Abutments

20
18
16
14

No. of posts
12
10
8
6
4
2
0
6 7 8 9 10 11 12 13
Post length (mm)

Fig 2 Number of posts according to post length as measured on the working cast; light gray
bars = lost posts and cores.
NO PART OF THIS ARTICLE MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM
PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.

The first 40 teeth were randomly assigned to one of 4, the reamers provided by the manufacturers were
the above groups. Radix-Anchor was excluded from used. The transition between the post and the core in
the randomization of the last 10 teeth, since there were all systems was conical, and the shape of the holes for
COPYRIGHT © 2003 BY QUINTESSENCE PUBLISHING CO, INC.
WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

indications that the system was not suitable for use in the posts was oval to prevent rotation, except in the
all situations. During preparation of the core buildup Radix-Anchor system. Care was taken to prepare the
on incisors with a large angulation between the crown outer surfaces of the root parallel at least 2 mm to cre-
and the root, a large portion of the core-retention part ate a ferrule effect. The finishing line was accomplished
of the Radix-Anchor post had to be removed. This cre- as a shallow chamfer preparation. A polyvinyl siloxane
ated a problem, as the teeth in the study had been ran- material, President (Coltène/Whaledent), was used for
domized, which made it mandatory to treat the se- impression taking. Before cementation of the post and
lected tooth with the randomized post-and-core core, the inner surface of the root canal was gently
combination. To preserve the random design of the coarsened with a medium-grit diamond on a low-speed
study as much as possible, group 4 was excluded contra-angle handpiece to create surface roughness.
from the randomization process and teeth were only The root canal was thoroughly cleaned with 3% hy-
assigned to groups 1, 2, or 3 (Fig 1). drogen peroxide, Tubulucid (an ethylenediaminete-
traacetic acid [EDTA]–containing cleaner; Dental
Clinical Preparation Therapeutics), and 96% alcohol using a rotating brush.
All posts and cores were cemented with zinc-phosphate
Five experienced dentists performed the initial cement (Dentsply/DeTrey). The crowns—metal ce-
prosthodontic treatment. Two clinicians placed one ramic or type III gold alloy with acrylic resin facing—
post and core each, one placed four posts and cores, were placed using the same luting cement.
and another placed five. The remaining 39 posts and The patients were re-examined regularly until 1999.
cores were inserted by one of the authors. The place- The final examination took place after between 72 and
ment of all posts and cores followed, if available, the 122 months (median 109 months). At the follow-up,
manufacturers’ recommendations concerning prepa- the cariologic and periodontal parameters were as-
ration, design, cementation, and choice of super- sessed for the experimental tooth and its contralat-
structure (acrylic resin–veneered gold alloy crowns or eral. If the latter was missing or included in the study,
metal-ceramic crowns). an adjacent tooth was used. The treatment was judged
Post space was prepared to accommodate a post of as successful if the post and core was still in place, with
at least 8 mm (Fig 2). Other objectives were that the no loss of retention after crown placement; exhibited
length of the post should equal the height of the artifi- no radiologic signs of fractures or sudden loss of bone
cial crown and that the apical part of the tooth should along the root surface indicating fracture or perforation;
contain at least 4 mm of gutta percha.21 In some cases, and exhibited no clinical signs of fracture such as pain,
these objectives were in conflict. The amount of gutta localized increased pocket depth, or fistulation.
percha was considered to be the most important fac-
tor. The diameters of the posts were at least 1.25 mm, Statistical Methods
with the exception of one Radix-Anchor size No. 1, di-
ameter 1.0 mm. A Davis reamer (JS Sjöding) was used The chi-square test was used to test whether differ-
to prepare the canals in group 1; the diameter of the post ences between the four groups of posts and cores
was measured in the coronal half. In groups 2, 3, and could be explained by mere chance or represented

Volume 16, Number 3, 2003 251 The International Journal of Prosthodontics

 Posts and Cores as Single-Crown Abutments Ellner et al

Table 1 No. of Teeth Failed and Failure Rate for real differences. The following levels of significance
Each Group Based on All Teeth were used: P ≤ .01 significant; .01 < P ≤ .05 almost
significant; and P > .05 not significant.
No. of teeth No. of teeth Failure
Group* total failed rate (%)
Results
1 14 0 0
2 13 1 8
3 13 0 0 The overall failure rate was 6% (Table 1). If group 4
4 10 2 20 (Radix-Anchor) is excluded—since the use of this
All 50 3 6 type of post and core was discontinued after the
1, 2, and 3 40 1 3
placement of the first 40—the overall failure rate was
*Group 1 = conventional cast tapered post and core, type III gold alloy; 3% at the final examination. It should be empha-
group 2 = ParaPost prefabricated post, No-Ox alloy, cast core type III
gold alloy; group 3 = ParaPost cast post and core, type III gold alloy; sized that until the first failure, 54 months after inser-
group 4 = Radix-Anchor titanium post, composite resin core. tion, all 50 original posts and cores had functioned sat-
isfactorily.
NO PART OF THIS ARTICLE MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM

One patient in group 2 died 58 months after place-

PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.

ment of a post and core; the restoration was func-

Table 2 Actual and Mean Functional Times (mo) for
Still-Functional and Failed Posts and Cores at Final tioning well at that time. Two posts and cores in
group 4, 9 mm long and placed on the maxillary left
COPYRIGHT © 2003 BY QUINTESSENCE PUBLISHING CO, INC.

Examination
WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

central incisor and second premolar, were lost after

Function Mean 54 and 79 months, respectively, because of loss of
Posts and cores* time time
retention. One post and core in group 3, 8 mm long
Group 1 and placed on the maxillary left canine, was lost be-
Posts still in place 81–122 101 cause of root fracture after 108 months. One patient
Group 2
Posts still in place 72–122 100 in group 1 was ill at the time of the final examination
Deceased patient 58 and could not attend, but reported that the post and
Group 3 core was in place and that there were no problems
Posts still in place 77–123 110
Root fracture 108 with it. Consequently, among the 30 patients who
Group 4 completed the study, 46 of 49 posts (94%) were still
Posts still in place 90–119 101 in place after a mean of 103 months.
Loss of retention 54, 79
At the final registration, there were no statistically
*See Table 1 for explantions of groups.
significant differences between the four groups. No
differences in function times between the four groups
were found at the final examination (Table 2). A de-
tailed analysis of the three failures did not indicate
any connection with factors such as maxillo-
mandibular relationships, occlusal interferences, un-
d e f balanced occlusion in lateral excursions or canine-
protected articulation, caries, or periodontal
affections. The conflicting objectives regarding post
length resulted in a post:crown ratio (c:d; Fig 3) of
< 1 for 34 of the 50 posts.
At the beginning of the study, periapical radiolu-
cencies were found around 10 roots. Three were still
a b c radiolucent at the final registration. Two of these
roots had a small radiolucency around excess gutta
percha material, and one root had a destruction 3 mm
in diameter. None of these defects changed in size
during the study period. Another three sites with no
signs of radiolucency at baseline had developed the
defect at the final registration. However, none of
these defects was larger than 1 mm in diameter or had
other symptoms that indicated a need for endodon-
Fig 3 Reference points for measurements: a = distance from tic retreatment or apical surgical therapy. The intra-
apex to marginal bone ridge; b = length of post protruding from
crown; c = total length of post; d = length of crown; e = length of oral radiographs of the other roots revealed no ab-
post and crown; f = total length of tooth. normal signs.

The International Journal of Prosthodontics 252 Volume 16, Number 3, 2003

 Ellner et al Posts and Cores as Single-Crown Abutments

Discussion the measurements described in Fig 3 were made and

used in the following calculations:
The results of two of the systems (groups 1 and 3) were
excellent—no failures. One of 13 failed in group 2, 1. c/d = length of post divided by length of crown
and two of 10 failed in group 4, but the differences be- 2. c/f = length of post divided by length of tooth
tween the four post-and-core systems were statistically 3. b/e = length of post protruding out of the crown
nonsignificant. The success rates exceeded (groups 1 divided by length of post and crown
and 3) or were comparable with (groups 2 and 4) those 4. a – (f – e) = length of post under bone margin
in previous retrospective studies. However, it should
be noted that the number of posts and cores in each No correlations between any of these calculations for
of the groups in the present study was comparatively the entire material and the calculations of the failed
small. Furthermore, during the collection of the pre- posts and cores could be found.
sent material, clinical experience indicated that the One special post-and-core system is sometimes
Radix-Anchor system seemed unsuitable for use on all touted as being superior to all others. The present
NO PART OF THIS ARTICLE MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM

teeth; its use was therefore discontinued after the first prospective long-term clinical study, covering up to
PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.

40 posts and cores had been placed. The compara- 10 years and comparing different systems, does not
tively negative results obtained with Radix-Anchor support such a dogmatic view. On the contrary, it can
should therefore be judged in this light. The reason for be concluded that—provided the procedures used in
COPYRIGHT © 2003 BY QUINTESSENCE PUBLISHING CO, INC.
WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

discontinuing use of the system was a fear of problems this study are strictly followed—the use of post-and-
with the retention of the core composite buildup. core buildups and abutments for single crowns has an
However, the reason for the two failures in group 4 excellent prognosis in the medium-term perspective.
was a loss of cement retention between the root and
the post. The crown had become loose in situ but did References
not fall off, as the post and core was retained by the
threads of the post. 1. Fernandes AS, Dessai GS. Factors affecting the fracture resistance
A pilot study covering a mean follow-up period of of post-core reconstructed teeth: A review. Int J Prosthodont 2001;
7.9 years16 analyzed the outcome of 112 post-and- 14:355–363.
2. Bergman B, Lundquist P, Sjögren U, Sundquist G. Restorative and
core buildups of a prefabricated metal post—Dentatus endodontic results after treatment with cast posts and cores. J
(Dentatus), Unimetric (Mailleter), or Radix-Anchor— Prosthet Dent 1989;61:10–15.
in combination with a composite core. The mean 3. Sorensen JA, Martinoff JT. Endodontically treated teeth as abut-
failure rate was 12.5%; Dentatus had the highest ments. J Prosthet Dent 1985;53:631–636.
4. Grieve AR, McAndrew R. A radiographic study of post-retained
(30%; n = 7), Radix-Anchor had the next highest
crowns in patients attending a dental hospital. Br Dent J 1993;174:
(13%; n = 6), and Unimetric had the lowest (2%; n = 197–201.
1). The mean follow-up times differed between the 5. Roberts DH. The failure of retainers in bridge prostheses—An
three systems; the follow-up time for Dentatus was the analysis of 2000 retainers. Br Dent J 1970;128:117–124.
longest. Furthermore, teeth with a poor initial prog- 6. Palmqvist S, Swartz B. Artificial crowns and fixed partial dentures
18 to 23 years after placement. Int J Prosthodont 1993;6:279–285.
nosis had been included in the material. These short-
7. Molin M, Bergman B, Ericson Å. A clinical evaluation of coni-
comings complicate a fair comparison between the cal crown retained dentures. J Prosthet Dent 1993;70:251–256.
three post-and-core systems, and the authors them- 8. Randow K, Glantz PO, Zöger B. Technical failures and some re-
selves underlined the need for controlled clinical tri- lated clinical complications in extensive fixed prosthodontics. An
als to verify the findings. For Radix-Anchor, the main epidemiological study of long-term clinical quality. Acta Odontol
Scand 1986;44:241–255.
cause of failure was dislodgment, and the authors
9. Laurell L, Lundgren D, Falk H, Hugoson A. Long-term progno-
recommended that this post be provided with an an- sis of extensive polyunit cantilevered fixed partial dentures. J
tirotation pin after insertion. It is interesting to note the Prosthet Dent 1991;66:545–552.
similarity in the mode of failure in this study and the 10. Lewis R, Smith B. A clinical survey of failed post retained crowns.
previous one.16 The Radix-Anchor post is cylindric Br Dent J 1988;165:95–97.
11. Weine FS, Wax AH, Wenckus CS. Retrospective study of tapered,
with one sharp thread, whose purpose is both to
smooth post systems in place for 10 years or more. J Endod 1991;
thread the root before luting and act as a retention el- 17:293–297.
ement after the luting procedure. The post is also 12. Mentink AGB, Meeuwissen R, Käyser AF, Mulder J. Survival rate
highly polished, possibly making retention of the ce- and failure characteristics of the all metal post and core restora-
ment more difficult. tion. J Oral Rehabil 1993;20:455–461.
13. Torbjörner A, Karlsson S, Ödman PA. Survival rate and failure char-
Several guidelines have been proposed to opti-
acteristics for two post designs. J Prosthet Dent 1995;73:439–444.
mize the success of a post,20–24 such as the length of 14. Sundh B, Ödman P. A study of fixed prosthodontics performed at
the post in absolute numbers and as related to tooth a university clinic 18 years after insertion. Int J Prosthodont 1997;
length, crown length, and so on. In the present study, 10:513–519.

Volume 16, Number 3, 2003 253 The International Journal of Prosthodontics

 Posts and Cores as Single-Crown Abutments Ellner et al

15. Ottl P, Laucer CH. Success rates for two different post and cores. 20. Zarb GA, Bergman B, Clayton JA, Mackay HF. The buildup of
J Oral Rehabil 1998;25:752–758. individual teeth. In: Prosthodontic Treatment for Partially
16. Mentink AGB, Creugers NHJ, Meeuwissen R, Leempoel PJB, Edentulous Patients. St Louis: Mosby, 1978:130–139.
Käyser A. Clinical performance of different post and core sys- 21. Morgano SM, Milot P. Clinical success of cast metal posts and
tems—Results of a pilot study. J Oral Rehabil 1993;20:577–584. cores. J Prosthet Dent 1993;70:11–16.
17. Lindhe LA. The use of composites as core material in root-filled 22. Glantz PO, Nilner K. Root canal posts—Some prosthodontic as-
teeth. II. Clinical investigation. Swed Dent J 1984;8:209–216. pects. Endod Dent Traumatol 1986;2:231–236.
18. Hatzikyriakos AH, Reisis GI, Tsingos N. A 3-year postoperative 23. Torbjörner A. Treatment management. Posts and cores. In:
clinical evaluation of posts and cores beneath existing crowns. Karlsson S, Nilner K, Dahl BL (eds). A Textbook of Fixed
J Prosthet Dent 1992;67:454–458. Prosthodontics. Stockholm: Gothia, 2000:173–186.
19. Petersson K, Petersson A, Olsson B, Håkansson J, Wennberg A. 24. Zmener O. Adaptation of threaded dowels to dentin. J Prosthet
Technical quality of root fillings in an adult Swedish population. Dent 1980;43:530–535.
Endod Dent Traumatol 1986;2:99–102.
NO PART OF THIS ARTICLE MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM

Literature Abstract
PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.

Comparative evaluation of casting retention using the ITI Solid abutment with
six cements.
COPYRIGHT © 2003 BY QUINTESSENCE PUBLISHING CO, INC.
WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

This study tested the retention of metal copings cemented on ITI Solid titanium abutments using
six different cements: zinc oxide (non-eugenol; Temp Bond NE); zinc oxide eugenol (IRM); zinc
phosphate (HY-Bond); resin-modified glass-ionomer (Protec Cem); zinc polycarboxylate
(Durelon); and resin cement (Panavia 21). Ten prefabricated burnout caps with wax loops at-
tached to the occlusal surface were cast with a noble alloy. The metal copings were cemented
onto solid titanium abutments connected to hollow-screw 3.8-mm ITI implants with a torque of 35
Ncm. A pullout test was carried out using a universal testing machine at a cross-head speed of
0.5 mm/min until decementation. The same 10 castings were used to test the six cements.
Based on the results, the retentive ability of the various cements was ranked in the following as-
cending order: Temp Bond, IRM, HY-Bond, Protec Cem, Durelon, and Panavia 21. The mean
loads at failure ranged from 3.18 kg (Temp Bond) to 36.53 kg (Panavia 21). The retentive value
of Temp Bond was not significantly different from IRM. HY-Bond and IRM provided statistically
similar retention values. Protec Cem and Durelon were not statistically different. Panavia 21
showed significantly higher retention than the other cements. The ability of the cements to retain
the castings differed from values obtained in studies where the same cements were used on nat-
ural teeth. The authors explained that this difference is likely due to the material and surface
characteristics of the implant abutment. Therefore, cement retention values reported in studies
that use teeth as abutments may not apply to cement-retained implant-supported restorations.

Mansour A, Ercoli C, Graser G, Tallents R, Moss M. Clin Oral Implants Res 2002;13:343–348.
References: 24. Reprints: Dr Carlo Ercoli, Division of Prosthodontics, University of Rochester Eastman,
Department of Dentistry, 625 Elmwood Avenue, Rochester, New York 14620. e-mail:
carlo_ercoli@urmc.rochester.edu—Swee-Chian Tan, Iowa City, Iowa

The International Journal of Prosthodontics 254 Volume 16, Number 3, 2003