The CEREC system allows the dentist to provide esthetic restorations in a single appointment.^1–3 The system includes an acquisition unit consisting of a portable computer, design software and an optical imaging system, and a milling chamber with two diamonds for milling the final restoration from prefabricated blocks of restorative material. The cavity preparation for an inlay, onlay, crown or veneer is recorded in the computer using an optical imaging procedure. The captured data are projected on the computer monitor of the acquisition unit, and software is used to design the contours of the final restoration. The software program interprets the design data to direct the diamonds in the milling chamber to mill the final restoration from prefabricated blanks of restorative material. The final restoration is adhesively cemented to the tooth to complete the restoration. This system offers a considerable time savings over conventional, laboratory-generated restorations that require multiple appointments.

Restorative material options for use with the CEREC system have been limited to Vita Mark II (Vita Zahnfabrik, Bad Säckingen, Germany), a fine-grained, feldspathic ceramic, or ProCAD (Ivoclar Vivadent, Amherst, N.Y.), a leucite-reinforced ceramic.⁴ A number of clinical studies have reported good performance and longevity with a direct placement CAD/CAM technique using blocks of ceramic material to fabricate the restorations.^5–8

A new resin-based composite material has been introduced for use in fabricating CAD/CAM restorations. Paradigm (3M ESPE) is a bisphenol A-diglycidyldimethacrylate/triethylene glycol dimethacrylate resin-based composite with filler composed of nanocrystalline zirconia in an amorphous silica matrix.⁹ The inorganic filler loading is 85 percent by weight with an average particle size of 0.6 micrometers, and it is radiopaque. However, there are no corresponding longitudinal studies documenting the clinical usefulness of CAD/CAM-generated resin-based composite inlays.

We conducted a longitudinal, randomized clinical trial to study the clinical performance of this recently developed resin-based composite material for CAD/CAM-generated adhesive inlays. The purpose of this clinical study was to evaluate the postoperative sensitivity associated with the adhesive luting technique and to evaluate the longitudinal clinical performance of ceramic and resin-based composite CAD/CAM inlays after three years of clinical service.

	SUBJECTS, METHODS AND MATERIALS

TOP ABSTRACT SUBJECTS, METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
All patients recruited for the study were fully informed of the nature of the study and the need for their long-term availability. Each patient signed a written consent form, approved by the University of Michigan Health Sciences Institutional Review Board, before being enrolled in the study. Each patient enrolled in the study had at least one two-surface (mesio-occlusal or disto-occlusal) or three-surface (mesial-occlusal-distal) carious lesion or defective restoration to be restored on a maxillary or mandibular premolar or molar. Each lesion or defective restoration exhibited sufficient size to extend at least one-half the intercuspal width of the tooth. All restored teeth were in functional occlusion. All restored teeth tested vital and were asymptomatic at the beginning of treatment. No more than two restorations were placed per patient.

We completed a caries risk assessment for each patient at baseline based on the number of restorations the patient reported having received in the previous 12 months. We scored patients as having a low caries risk if they had had zero restorations or one restoration placed in the previous 12 months; as having a moderate caries risk if they had had two or three restorations placed in the previous 12 months; and as having a high caries risk if they had had four or more restorations placed in the previous 12 months. We scored all enrolled patients as having a low caries risk. Exclusion criteria included nonvital teeth, sensitive teeth, teeth with prior endodontic treatment, teeth with a history of direct or indirect pulp capping procedures, and patients with significant untreated dental disease (including periodontitis and rampant caries).

Two clinicians (K.L. and D.J.F.), each with more than five years of experience in the CEREC technique, placed 40 resin-based composite inlays of Paradigm and 40 porcelain inlays of Vita Mark II for a total of 80 inlays. One operator (K.L.) placed 62 of the inlays, and the second operator (D.J.F.) placed the remaining 18 inlays. The clinician determined the preoperative shade of the tooth to be restored using a shade guide (Vita Classic, Vita Zahnfabrik) before starting the restorative treatment. After administering local anesthetic, the clinician used rubber dam isolation for cavity preparation, optical imaging and adhesive cementation for every restoration. The clinician removed the rubber dam after adhesively cementing the restoration and completed the final occlusal adjustment and polishing.

Inlay cavity preparation consisted of butt joint margins without bevels. All walls were tapered 6 to 8 degrees from the pulpal floor to the cavosurface margin. No bases or liners were used in the study. The flare of the proximal boxes conformed to standard criteria for an inlay with the proximal margins exposed for convenience in finishing. The clinicians adhered strictly to the manufacturer’s instructions in the imaging, computer design and machining of the restorations. They used a CEREC 2 unit with Operating System 1.21 to design and mill the inlays and the extended machining option for the milling process.

After completing the computerized graphic design of the inlay, the operator opened the envelope containing the random assignment of the prefabricated block to be used for the specific restoration. Delaying the random assignment of restorative material to this point in the treatment process ensured that shade determination, cavity preparation and design were not biased by the choice of restorative material.

The clinicians milled 40 inlays from prefabricated ceramic blocks of Vita Mark II. After trial seating, the clinicians etched the internal surfaces of the ceramic inlays for 60 seconds with 4.9 percent hydrofluoric acid gel, rinsed them for 20 seconds, and then air dried them with oil-free air. They applied prehydrolyzed silane coupling agent (RelyX Ceramic Primer, 3M ESPE) to the etched restoration before cementing it.

The clinicians then milled 40 inlays from prefabricated composite blocks of Paradigm. After trial seating, they air-abraded internal surfaces with 50-µm aluminum oxide at 40 pounds per square inch. They applied a single layer of Single Bond Adhesive (3M ESPE) to the inlay and cured it for 20 seconds. They acid-etched cavity preparations for 30 seconds with 37 percent phosphoric acid, and then rinsed them for 20 seconds. They dried the tooth lightly with high-volume evacuation and blotted dry any pooled water with cotton to ensure a moist surface and to avoid dehydration of the cavity preparation. The clinicians applied Single Bond (3M ESPE) bonding system to all prepared tooth surfaces before seating the restoration and cured it for 10 seconds. They used RelyX Adhesive Resin Cement (ARC) (3M ESPE) for all inlays and cured it with a visible–light-curing unit for 40 seconds from the facial, lingual and occlusal directions for a total curing time of two minutes. They used a series of diamond finishing burs, rubber abrasive points and cups, finishing strips and diamond polishing pastes for removing excess cement, adjusting the occlusion and completing final polishing.

The clinicians took intraoral digital photographs at baseline to document the preoperative condition, cavity preparation, inlay try-in and postoperative conditions. They made a postcementation quadrant impression of each test restoration in a polyvinylsiloxane material and poured casts in an epoxy die material for indirect evaluation.

Data collection. To evaluate the immediate postoperative sensitivity, the clinicians contacted patients by telephone once a week after the initial appointment. We used the telephone interview as a follow-up procedure to minimize recall loss as the patient was not required to return to the clinic. During the telephone interview, we made a criterion-referenced rating of tooth sensitivity (Box 1).

	RESULTS

TOP ABSTRACT SUBJECTS, METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSIONS REFERENCES

View larger version (111K): [in this window] [in a new window]   Figure 1. Vita Mark II (Vita Zahnfabrik, Bad Säckingen, Germany) inlay at (first row, left to right) six months, one year, (second row, left to right) two years and three years.

View larger version (109K): [in this window] [in a new window]   Figure 2. Paradigm (3M ESPE, St. Paul, Minn.) inlay at (first row, left to right) six months, one year, (second row, left to right) two years and three years.

Clinical evaluation. The ratings for anatomical form, margin finish, surface finish, tooth fracture and recurrent caries for both groups remained essentially unchanged from baseline to three years with all ratings in excess of 90 percent Alfa (Table 2, page 1719).

View larger version (89K): [in this window] [in a new window]   Figure 3. Vita Mark II (Vita Zahnfabrik, Bad Säckingen, Germany) inlay with chipped margin ridge at six months.

View larger version (88K): [in this window] [in a new window]   Figure 4. Vita Mark II (Vita Zahnfabrik, Bad Säckingen, Germany) inlay fractured at three years, requiring replacement.

At baseline and six months, we rated 100 percent of the resin-based composite and porcelain inlays Alfa for marginal discoloration (Table 2). There was no significant difference in marginal discoloration between the two materials at three years (² test, P = .945), with 83.8 percent of the resin-based composite inlays and 91.2 percent of the porcelain inlays rated Alfa.

At baseline, 100 percent of resin-based composite inlays and 85 percent of the porcelain inlays were rated Alfa for color match (Table 2). At three years, we rated 86.5 percent of the resin-based composite inlays and 58.8 percent of the porcelain inlays Alfa for color match. The resin-based composite inlays had a significantly better color match than did the porcelain inlays at three years (² test, P < .05). The resin-based composite inlays had no significant difference in color match at any recall period compared with baseline (Wilcoxon SRT, P < .05). The porcelain inlays had a significant decrease in color match at six months compared with baseline; however, there was no significant difference noted between six months and three years (Wilcoxon SRT, P < .05).

	DISCUSSION

TOP ABSTRACT SUBJECTS, METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Patients’ experiencing postoperative sensitivity after undergoing adhesive restorative procedures is not an uncommon problem. Although there are a number of reports documenting postoperative sensitivity for direct or indirect adhesive restorations, few of them deal directly with CEREC restorations. Sjögren and colleagues¹¹ reported that 10 of 72 patients had postoperative sensitivity with Vita Zahnfabrik’s Vita Mark I or II ceramic inlays. However, Heymann and colleagues⁶ reported no postoperative sensitivity at any recall interval in their four-year clinical trial of CEREC ceramic inlays. Fasbinder and colleagues¹² reported that 13 percent of 92 Vita Mark II onlays were rated slightly sensitive at one week and 4 percent at two weeks. They found that all sensitivity was resolved by one month, and that there essentially was no postoperative sensitivity throughout the remainder of the study.

We found a similar minimal amount of postoperative sensitivity in our study. All sensitivity resolved by two weeks and was not a factor during the three-year study period. There are several possible reasons for the lack of postoperative sensitivity in this study. The use of rubber dam isolation for control of the operating field for all test restorations throughout the preparation, design and cementation steps ensured a clean, isolated tooth surface for adhesive bonding. The CAD/CAM technique also may play a role in minimizing postoperative sensitivity. The ability to deliver the porcelain and resin-based composite inlays in a single appointment prevented the potential for tooth contamination during the temporization phase. Also, the use of manufactured blocks of porcelain and resin-based composite minimized the influence of polymerization shrinkage since it was limited to the thickness of the resin cement.

The failure rate of Vita Mark II porcelain inlays made with the CEREC technique has been reported to be low. Sjögren and colleagues⁸ reported four fractures of 66 Vita Mark II inlays during a five-year period. Pallesen and van Dijken¹³ reported a single fractured Vita Mark II inlay in 16 pairs of CAD/CAM inlays during an eight-year period. Berg and Derand¹⁴ reported three fractures in 115 Vita Mark II inlays during a five-year period. In a systematic review of clinical studies of intracoronal CEREC restorations Martin and Jedynakiewicz¹⁵ reported a mean survival rate of 97.4 percent during a four-year period. The primary reasons for failure were reported as fracture of the ceramic, fracture of the supporting tooth and postoperative sensitivity. Fracture of the ceramic was generally a result of occlusal stress or insufficient ceramic thickness. Otto and De Nisco⁷ reported an 8 percent failure rate for ceramic inlays after 10 years of clinical service. Of the failures, 53 percent were caused by ceramic fracture and 20 percent by tooth fracture. Hickel and Manhart¹⁶ reviewed clinical studies in the dental literature during the 1990s and reported annual failure rates of posterior restorations in stress-bearing areas as 0 percent to 11.8 percent for composite inlays, 0 percent to 7.5 percent for ceramic inlays and 0 percent to 4.4 percent for CAD/CAM ceramic restorations. They also reported bulk fracture as a frequent cause of failure for ceramic inlays. The results of our study also reflect low failure rates for the CEREC inlays. Fracture of the porcelain inlays was the primary failure mechanism for the Vita Mark II material, as four of the 40 Vita Mark II inlays fractured. Only one of the fractured inlays required treatment. Two of the inlays had occlusal isthmus fractures that had no clinical symptoms, as the restorations remained intact and bonded in place. Of note is that none of the Paradigm inlays were chipped or fractured during the study. It will be interesting to see if the fracture resistance of the resin-based composite inlays will be maintained over longer recall intervals.

The decrease in rating for color match was more a function of the modification of the tooth color over time rather than a discoloration of the porcelain inlays.

The porcelain and resin-based composite inlays had no significant changes in surface finish and anatomical form, a finding that is consistent with other reported clinical studies^6,8,13 (Figures 1, 2, 5 and 6). Gladys and colleagues¹⁷ reported a three-year clinical evaluation of CAD/CAM ceramic inlays and composite inlays (P-50, 3M ESPE). The surfaces of 89 percent of the inlays were rated smooth, and all restorations were rated clinically acceptable. Although there was an overall 71 percent excellent color match, 50 percent of the composite inlays were evaluated as dull. Thordrup and colleagues¹⁸ reported a five-year study comparing CEREC porcelain inlays with indirect and direct composite restorations, as well as an indirect porcelain inlay. They reported that the indirect composite inlays had a rougher surface texture—which they attributed to the larger filler particle size—than the direct composite and porcelain materials. There were localized rough areas on the remaining inlays attributed to wear at functional cusp locations. In our study, both the Vita Mark II and Paradigm materials had surface properties that maintained the surface finish and anatomy over time.

View larger version (93K): [in this window] [in a new window]   Figure 5. Vita Mark II (Vita Zahnfabrik, Bad Säckingen, Germany) distal-occlusal inlay with evidence of margin detection (Alfa-2) at two years.

View larger version (81K): [in this window] [in a new window]   Figure 6. Paradigm (3M ESPE, St. Paul Minn.) mesial-occlusal-distal inlay with evidence of margin detection (Alfa-2) at two years.

Margin adaptation often is discussed as a critical factor in the longevity of indirect restorations. It has been suggested that an increase in margin gap size may lead to degradation of the adhesive bond, in turn resulting in microleakage and recurrent caries. However, the consequences of these potential changes are not well-documented over the long term. In our study, margins were detectable clinically for both materials as early as six months. There were no significant differences in the margin adaptation between the porcelain and resin-based composite inlays after three years, as both materials had 60 to 65 percent undetectable margins. At the one-year recall, there was a significant difference in the margin adaptation with the resin-based composite inlays having a greater percentage of nondetectable margins (91.4 percent) compared with the porcelain inlays (75.7 percent). This was consistent with a study by Gladys and colleagues¹⁷ in which they reported no significant difference in margin adaptation between Vita Mark I porcelain and P-50 resin-based composite inlays after three years, but with margin detection noted as early as six months. The P-50 material had the best inlay-lute margin interface at three years. Pallesen and Qvist²⁰ also reported an equivalent wear of the resin cement and the resin inlays over 11 years. They evaluated resin-based composite fillings and inlays with modified PHS criteria, and they did not identify any significant margin wear. However, examination of the stone dies made at each recall interval revealed wear of the resin luting agent along the enamel occlusal margins in more than one-half of the models. This finding illustrates the inherent problem of detecting measurable amounts of margin wear during a clinical examination. In this study, the modified U.S. PHS criteria were refined to try and detect subtler margin changes. The modified criteria distinguished between margins that were nondetectable, were detectable in greater or less than 50 percent of the occlusal margin but less than 1 mm in depth, and detectable with a crevice formation deeper than 1 mm. Evaluation of the Alfa-1 (nondetectable) and Alfa-2 (detectable less than 50 percent of the margin and less than 1 mm deep) scores revealed a significant increase in margin detection for the porcelain inlays at the one-year recall and for the resin-based composite inlays at the two-year recall (Figures 5 and 6). By the three-year recall, there was no significant difference in margin detection between the composite and porcelain inlays. Zuellig-Singer and Bryant²¹ evaluated the margin adaptation of Vita Mark II inlays with a scanning electron microscope. They reported a significant decrease in the percentage of continuous margins at the luting agent-porcelain interface. They found no significant change in the luting agent-enamel interface over three years of clinical service. The detectable area is the luting agent-porcelain interface. It seems logical from these studies that the resin-based composite inlay has a wear rate similar to that of the resin-based composite luting agent, thus masking the initial wear of the inlay margin. It is not until sufficient wear at the margin occurs, exposing the enamel margin, that it can be detected similar to that of the porcelain. We plan a quantitative evaluation of the wear with an indirect technique at the five-year recall. In spite of the detected margin wear, we noted no significant margin discoloration or any secondary caries for either material over the three years. This would indicate that the margin wear is an occlusal surface phenomenon and was not accompanied by a breakdown in the adhesive bond to the tooth.

There were no significant differences in the margin adaptation between the porcelain and resin-based composite inlays after three years, as both materials had 60 to 65 percent undetectable margins.

	CONCLUSIONS

TOP ABSTRACT SUBJECTS, METHODS AND MATERIALS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Postoperative sensitivity was not a significant finding from baseline to three years for either material. Initial color match for both materials was rated very good and was maintained better by the resin-based composite inlays at three years. Tooth color match of the porcelain inlays decreased by the six-month recall, but then remained unchanged to the three-year recall. Margin adaptation initially was very good for both materials, with an increase in margin discontinuity caused by apparent wear of the composite luting agent. The resin-based composite inlays performed equally as well as the porcelain inlays at three years in all categories, with clinical advantages noted in fracture resistance and better color match to the tooth.