Purpose. The purpose of this clinical study was to assess the restoration quality of and gingival response to CAD/CAM
fabricated posterior single-tooth restorations with different processing technologies.
Material and methods. Twenty-two individuals in need of posterior complete coverage crowns were recruited under an
institutional review board approved protocol. Teeth were randomized to 1 of 3 groups: metal ceramic, lithium disilicate, and
monolithic zirconia. An unprepared or minimally restored tooth on the contralateral side was chosen as a control tooth
for gingival measurements with each participant. Teeth were prepared and scanned intraorally by 1 of 3 experienced
practitioners. A total of 32 restorations were digitally designed and fabricated with either milling technology or
rapid-prototype printing and casting with conventional porcelain application. Restorations were evaluated with modified
United States Public Health Service criteria for contour, marginal adaptation, occlusion, and shade. Gingival crevicular fluid
volume and bleeding on probing were recorded preoperatively, at 1-month and 6-month postcementation visits. Polyvinyl
siloxane impressions were made of the buccal margin of cemented restorations and evaluated with microcomputed
tomography to assess marginal adaptation (horizontal discrepancy). The Mantel Haenszel row mean score was used to assess
whether the crown systems differed with respect to the modified United States Public Health Service criteria. Linear mixed
models were used to assess whether the average gingival volumes were affected by the explanatory variables (crown system,
tooth status [treated vs control], or visit). A generalized estimating equation approach was used to assess whether
bleeding on probing was affected by the explanatory variables. One-way ANOVA was used to assess marginal discrepancy
values among the crown systems (a¼.05 for all tests).
Results. Twelve metal ceramic, 10 lithium disilicate, and 10 zirconia restorations were fabricated for 22 participants. Zirconia
restorations were significantly different from the other 2 crown systems (P<.001) with respect to occlusion. No occlusal
adjustment was needed on 80% of the zirconia restorations. The average gingival crevicular fluid volumes did not differ among
crown systems, between treated and control groups, or over time. The average horizontal marginal discrepancy was
significantly different between lithium disilicate and zirconia crowns (P¼.027), with zirconia crowns having the least amount
of horizontal marginal discrepancy.
Conclusions. Given the small sample size and limitations of this study, CAD/CAM-generated restorations for posterior teeth
made from different materials had acceptable clinical results. (J Prosthet Dent 2014;112:770-777)
This study was supported by an American Academy of Fixed Prosthodontics Stanley D. Tylman grant.
a
Former Resident, Department of Prosthodontics, University of North Carolina at Chapel Hill School of Dentistry, Private practice,
Colorado Springs, Colo.
b
Director and Stallings Distinguished Professor, Graduate Prosthodontics, University of North Carolina at Chapel Hill School of Dentistry.
c
Clinical Assistant Professor, Department of Prosthodontics, University of North Carolina at Chapel Hill School of Dentistry.
d
Clinical Assistant Professor, Department of Prosthodontics, University of North Carolina at Chapel Hill School of Dentistry.
Clinical Implications
Practitioners have multiple options for generating single-tooth
restorations with computer-aided design and computer-aided
manufacturing technology. Irrespective of available material choices,
a digital workflow that involves intraoral scanning and digital design
may provide a clinically acceptable single-tooth restoration.
Computer-aided design and computer- The direct intraoral capture of a pre- (USPHS) criteria. Secondary outcomes
aided manufacturing (CAD/CAM) pared abutment found more accuracy included marginal adaptation measured
dentistry has evolved dramatically in than the scanned polyether impression by micro–computed tomography (mi-
the past 30 years. Most recently, or gypsum cast. Average marginal dis- cro-CT) analysis and the gingival
numerous CAD/CAM systems with crepancies of CAD/CAM restorations response to CAD/CAM crown restora-
broad capabilities that range from have been reported, which ranged tions evaluated at the level of inflam-
implant restoration design and fabri- from 35 to 71 mm, and clinical studies mation with gingival crevicular fluid
cation to orthodontic appliance plan- with scanning electron microscopy (GCF) volume and bleeding on probing
ning and manufacture have become analysis have equivalent values.5,8-10 (BOP). The null hypotheses were that no
widely available for clinicians and A clinical study by Brawek et al11 re- difference would be found in crown
dental laboratories. In addition, de- ported a mean marginal discrepancy quality among metal ceramic (MC), LD,
mand for esthetic materials has in- value of 51 mm for veneered Zr crowns and Zr crowns fabricated with CAD/
creased1 in restorative dentistry, with fabricated with intraoral scanning CAM technologies, no difference would
concurrent advances in materials sci- techniques and digital fabrication. be found in gingival response among the
ence that provide new monolithic Even with studies that found com- crown systems, and no difference would
materials that synergize with CAD/ parable marginal adaptation with be found in the marginal discrepancy
CAM technology. Two materials, lith- standard metal ceramic restorations, values between crown systems and
ium disilicate (LD) and monolithic the longevity of ceramic materials are intraoral scanners.
zirconia (Zr), have become popular, of concern, especially those with
and both are fabricated with CAD/ veneering methods for improved es- MATERIAL AND METHODS
CAM systems.2 A recent survey of thetics. A 94.8% 8-year survival rate
laboratory fabrication projections for was reported by Gehrt et al12 for 94 Patients of the University of North
restorative materials estimates that, by single-unit fixed dental prostheses for Carolina (UNC) at Chapel Hill School
the year 2017, ceramic materials will veneered LD restorations with a 5.5% of Dentistry were screened and enrolled
be used to fabricate approximately rate of chipping or fracturing of res- in this study under institutional review
42% of fixed dental restorations.1 torations. Fasbinder et al13 reported a board approval (UNC institutional re-
The adoption of digital manufacture 100% survival rate of 62 LD crown view board 11-2099). Before treatment
for crowns involves the professional restorations in 43 individuals over a began, radiographs, tooth shade, GCF
assessment of all attributes of the res- 2-year recall period. No incidence of volume, and BOP measurements were
toration. Notably, attention has been crown fracture or chipping was re- recorded for all treated teeth as well
focused on the fit and marginal ad- ported nor relatively high alpha scores as a minimally restored or unprepared
aptation of CAD/CAM restorations.3 for color and marginal adaptation. control tooth on the contralateral side.
As discussed in a recent review by Higher chipping rates of 15% to 25% Shade was matched to a Vita Classic
Miyazaki et al,4 crown margins can have been reported for veneered Zr shade guide for treated teeth. GCF
be difficult to capture with intraoral restorations.14,15 A monolithic resto- volume was collected on the buccal and
scanning, not only because of their ration would appear to solve some of lingual surfaces of selected and control
design, but their proximity to gingival the technical complications associated teeth with crevicular fluid strips (Perio-
tissues, adjacent teeth, and sulcular with ceramic materials, despite limited paper; Oraflow) and a Periotron 8000
fluids. Other studies have concluded clinical data. (Oraflow) according to the manufac-
no significant difference among intrao- This prospective study compared turer’s instructions. BOP was measured
ral scanning, a scan of an impression, different CAD/CAM crown procedures. with a periodontal probe (UNC-15
or a gypsum cast.5,6 Güth et al7 The primary outcome measurement was Periodontal Probe; Hu-Friedy) on the
examined the accuracy of digital restoration quality assessed with modi- treated and control teeth. The crown
models with an in vitro experiment. fied United States Public Health Service system was randomized to treated
Batson et al
772 Volume 112 Issue 4
teeth by using computerized software the iTero scanner. Once scanning was stains were added for characterization
(Random Allocation Software; http:// complete, an interim restoration was (Empress stains; Ivoclar Vivadent).
mahmoodsaghaei.tripod.com/Softwares/ fabricated from a bis-acryl material The LD crowns were fabricated
randalloc.html). A total of 32 crowns (Integrity; Dentsply Caulk), polished, with the CAD software within the E4D
were fabricated for 22 participants. and cemented with interim cement system. The restorations were sent
Teeth that required a foundation (Temp-Bond NE; Kerr Dental). electronically to the E4D mill, and LD
for appropriate resistance or retention Scans for MC and Zr crowns were blocks (eMax CAD; Ivoclar Vivadent)
form were treated with either an sent electronically to Align Technology were milled in the selected shade. The
amalgam core (Sybralloy; Kerr Dental) Inc. Preparation margins were marked E4D design software die spacing
or composite resin core (Comp-Core; electronically by the clinician when sig- settings of 0.10 mm were used. After
Premier Dental). All the teeth were pre- nificant deviations were noted from the milling, sintering was completed ac-
pared by 1 of 3 experienced practitioners default margin. Solid casts were ordered cording to the manufacturer’s di-
(E.B., L.C., I.D.) by using standard rec- for all MC and Zr restorations (Cadent rections. Characterization and glazing
ommended preparation guidelines of model; Align Technology Inc). One were completed with appropriate stains
1.5 to 2.0 mm occlusal reduction, and dental laboratory fabricated all the MC (IPS eMax Ceram; Ivoclar Vivadent).
1.0 to 1.5-mm axial reduction with a and Zr crowns. Both the MC and Zr One clinician (E. B.) recorded the
deep chamfer margin circumferentially. crowns were designed with digital design modified USPHS crown-quality criteria
Margin placement was designated at no software (3Shape software; 3Shape). as listed in Table I during the crown
more than 0.5 mm subgingivally. After Die-space allowances were set within the delivery appointment. All crowns were
preparation, the teeth were isolated design software at 0.030 mm for Zr fitted by first verifying the interproximal
from the gingival tissues by placing crowns, and 0.040 mm for MC crowns contacts and then verifying the fit with
appropriately sized plain displacement according to the type of material and the addition silicone (black for ceramic,
cord (Ultrapak; Ultradent Products Inc) manufacturer’s recommendations. MC white for MC; Fit-Checker; GC America
and were scanned according to the type copings were produced by using rapid Inc) to disclose and verify the fit of
of restoration assigned; MC and Zr res- prototype printing (3-dimensional the intaglio surface. Necessary internal
torations were assigned to the iTero printing) technology and an appro- adjustments were made for MC crowns
scanner (Align Technology, Inc), and priate polymer (Envisiontec Ultra 2 3-D with a carbide bur (no. 4 round bur;
LD crowns were assigned to the E4D printer and EC 1000 Photopolymer; Hu-Friedy). For Zr and LD crowns, ad-
scanner (Planmeca-E4D) linked to the EnvisionTEC). Printed copings were justments were made on the prepared
E4D milling unit. All the teeth were invested and cast from a high noble alloy tooth surfaces when possible with
scanned by using the recommended se- followed by the application of porcelain a water-cooled fine diamond rotary in-
quences or patterns as determined by (Argedent Euro; The Argen Corporation strument (Two-Striper Diamond; Pre-
each scanner. Scanned preparations and IPS d.Sign, In-Line Porcelain; mier Dental). Marginal fit was verified
were reviewed chairside, and, if neces- Ivoclar-Vivadent). The Zr restorations by explorer feel. Occlusion was evalu-
sary, adjustments were made and the were fabricated by using milling tech- ated with occlusal indicating paper
tooth rescanned. An intraocclusal re- nology (Wieland Mini-mill; Wieland (Accufilm; Parkell Dental), and any
cord was made with an occlusal regis- Dental). Intrinsically colored monolithic necessary adjustments were made with
tration material for teeth assigned to the Zr blocks (Zenostar; Wieland Dental) polishing wheels (Dialite polishers;
LD group (Virtual CADbite; Ivoclar were milled in the green state and then Brasseler Dental). If extensive adjust-
Vivadent). MC and Zr crown intra- sintered according to manufacturer ment was necessary, the crown was
occlusal scans were made as directed by recommendations. If necessary, extrinsic reglazed. Once seated, the shade of the
Table I. Modified United States Public Health Service criteria for crown evaluation
Marginal
Results Adaptation Crown Contour Shade Occlusion
S: Acceptable Less than ideal Less than ideal but minimal Less than ideal but no Less than ideal but no
but acceptable or no changes required changes required changes required/minimal
adjustments necessary
T: Acceptable/modifications Less than ideal, Additions or Staining or other shade Adjustments necessary
needed adjust or remake reductions necessary modifications required
Marginal adaptation
MC 3 0 8 1 12
Zr 1 0 5 4 10
LD 2 1 7 0 10
Total 6 1 20 5 32
Contour
MC 0 2 7 3 12
Zr 0 0 9 1 10
LD 0 3 7 0 10
Total 0 5 23 4 32
Shade
MC 0 2 6 4 12
Zr 0 6 3 1 10
LD 0 1 9 0 10
Total 0 9 18 5 32
Occlusion
MC 0 0 7 5 12
a
Zr 0 0 2 8 10
LD 0 3 7 0 10
Total 0 3 16 13 32
For USPHS Shade, P¼.06; for USPHS Contour, P¼.16; for USPHS Margins, P¼.07; for USPHS Occlusion, P¼.0005.
USPHS, United States Public Health Service; MC, metal ceramic; Zr, zirconia; LD, lithium disilicate.
a
Denotes statistically significant values assessed by Mantel Haenszel row mean score statistic.
Batson et al
774 Volume 112 Issue 4
structure and compound symmetry 1-month recall visit, and 13 participants Tables III and IV. The P values for all
in the generalized estimating equation did not return for the 6-month recall explanatory variables associated with
method. One-way ANOVA was used to visit. Twenty-nine crowns were evaluated the gingival volumes and BOP are pro-
determine whether the average hori- at the 1-month recall visit, and 19 vided in Table IV.
zontal marginal discrepancies differed crowns at the 6-month recall visit. The average horizontal marginal
among the crown systems. For those No statistically significant associa- discrepancy as determined by micro-CT
that were significant, pairwise compar- tion was found among crown types and analysis was statistically significantly
isons were used between the crown their marginal adaptation, shade, or different among the crown systems
systems and scanners. The Bonferroni contour with the Mantel Haenszel row (P¼.003). The number of crowns per
method was used to obtain adjusted P mean score statistic. The majority group and the type of horizontal
values (a¼.05). of restorations were scored in the discrepancy as determined by micro-CT
“acceptable” category with the USPHS analysis are displayed in Figure 1. A
RESULTS criteria. The distribution of crowns study LD crown as seen by micro-CT
graded within each USPHS criterion are imaging is demonstrated in Figure 2.
Six crowns (3 MC, 2 LD, 1 Zr) were presented in Table II. Zr crowns were Horizontal marginal discrepancy mea-
rejected for unacceptable marginal ad- statistically significantly different from surements made with micro-CT imag-
aptation and were remade. Thus, the LD and MC crowns for occlusion ing are depicted in Figure 3. Pairwise
remake rate because of clinically un- (P<.001). Eighty percent of Zr crowns comparisons were used to determine
acceptable marginal adaptation was had an “excellent” rating and needed which systems differed. LD and Zr were
18.8%. These 6 crowns were among the no occlusal adjustment. Fifty-eight significantly different (P¼.027), with
first 12 fabricated, defining a learning percent of MC restorations and 100% LD crowns having larger values of hor-
curve related to the software designa- of LD restorations needed occlusal izontal discrepancies. Descriptive sta-
tion of margin location in design. Of adjustment before insertion. No statis- tistics as well as pairwise comparisons
these six, 2 MC crowns were remade tically significant differences were found for horizontal discrepancy values are
with conventional impression and lab- among the 3 crown systems for GCF given in Tables V and VI.
oratory techniques, and were excluded volumes or BOP. GCF volumes and
from the horizontal marginal discrep- BOP were not significantly different DISCUSSION
ancy analyses. The remaining 4 crowns between the treated and the control
that were remade were fabricated from teeth or between visits. Descriptive sta- Although the crown systems did not
new intraoral scans and used the same tistics and P values for all explanatory differ with respect to shade, contour,
CAD/CAM processes for fabrication. variables associated with the gingival and marginal adaptation, Zr crowns
Three participants did not return for the volumes and BOP are provided in were superior in crown occlusion. Few Zr
Table III. Descriptive statistics for gingival parameters and treated teeth (N¼32)
Visita
Crown 1-mo 6-mo
Gingival Parameter System Pretreatmentb Postcementationc Postcementationd
Buccal surface GCF volume, mean (SD) MC 46.25 23.95 43.75 14.06 32.67 14.51
Zr 47.80 22.79 39.00 12.12 39.29 14.40
LD 32.00 9.56 44.13 24.51 43.33 19.10
Lingual surface GCF volume, mean (SD) MC 33.25 12.17 43.17 17.48 37.50 20.05
Zr 44.10 22.49 37.00 17.61 42.57 23.73
LD 39.40 16.77 56.88 23.74 39.83 15.72
Bleeding on probing, yes/no (% of yes) MC 8/4 (67) 5/7 (42) 3/3 (50)
Zr 6/4 (60) 5/4 (56) 2/5 (29)
LD 5/5 (50) 2/6 (25) 2/4 (33)
MC, metal ceramic; SD, standard deviation; GCF, gingival crevicular fluid; Zr, zirconia; LD, lithium disilicate.
a
MC (n¼12 [missing¼0] at visit 0, n¼12 [missing¼0] at visit 2, n¼6 [missing¼6] at visit 3); Zr (n¼10 [missing¼0] at visit 0, n¼9 [missing¼1] at visit 2,
n¼7 [missing¼3] at visit 3); LD (n¼10 [missing¼0] at visit 0, n¼8 [missing¼2] at visit 2, n¼6 [missing¼4] at visit 3).
b
Visit 0.
c
Visit 2.
d
Visit 3.
CONCLUSIONS
Batson et al