B Van Meerbeek * M Vargas * S Inoue 8 Y Yoshida * M Peumans * P Lambrechts * G VanHerle
Operative Dentistry Supplement 6, 2001, 119-144
During the last three decades clinicians have been confronted with a continuous and fairly rapid turnover in adhesive materials. It started in the mid-'60s with the advent of the first commercialised restorative resin composites, followed in the early '70s with the introduction of the acid-etch technique in clinical practice. Since then, there has been ongoing progress in developing more refined and diversified restorative composites along with the production of steadily improved bonding agents. Effective adhesion to enamel has been achieved with relative ease and has repeatedly proven to be a durable and reliable clinical procedure for routine applications in modern adhesive restorative dentistry. Although adhesion to dentin is not as reliable as adhesion to enamel, today's adhesives produce superior results in laboratories (Perdigão & Lopes, 199; Inoue & others, 2000a,b; Tanumiharja, Burrow & Tyas, 2000), along with improved clinical effectiveness (Van Meerbeek & others, 1994b, 1996, 1998a; Brunton & others, 1999; Folwaczny & others, 2000; Tyas, 2000; Van Dijken, 2000), thereby, approaching enamel-bonding performance.
Early one-step dentin bonding agents became multi-step systems with more complicated, time-consuming and technique-sensitive application procedures. In the early '90s, the selective enamel-etching technique was replaced by a total-etch concept. Since then, universal enamel-dentin conditioners have been simultaneously applied to enamel and dentin. Now that today's total-etch adhesives have reached a clinically acceptable bonding effectiveness, most recent research and development efforts have focused on simplifying the multi-step bonding process and reducing its sensitivity to errors of inaccurate or incorrect clinical handling (Sano & others, 1998; Finger & Balkenhol, 1999; Inoue & others, 2000b).
In addition to interposition of a resin-based adhesive system between the restorative material and the remaining tooth structure, bonding to tooth tissue can also be clinically achieved directly using glass ionomer cements (Davidson & Mjör, 1999). Glass ionomer-based materials have an auto-adhesive capacity thanks to their specific chemical formula and structural nature (Wilson, Prosser & Powis, 1983; Van Meerbeek & others, 1998b; Yoshida & others, 2000). Parallel with the progress made in resin-based adhesives, glass ionomer technology has undergone many improvements and modifications of its original chemistry since being developed in the early '80s by Wilson & Kent (1971). A thorough discussion of self-adhering glass ionomer materials is beyond the certainly the case for three out of four "one-bottle" adhesives tested but also for the strong self-etch adhesive NRC/Prime&Bond NT and even the mild self-etch adhesive Clearfil SE Bond, of which the mTBS does not significantly differ from that of Optibond FL. Some pre-testing failures were recorded for the experimental PQ/Universal (three out of 10 specimens could be tested) when used following a self-etch approach, for One-up Bond F (six out of 11 were tested) and for Unifil Bond (10 out of 11 were tested). All adhesives in this study were bonded to enamel on which a 600-grit smear layer was prepared beforehand. In this respect, another study by Kanemura & others (1999) revealed that two other self-etch adhesives (Clearfil Liner Bond 2, Kuraray; Mac Bond 2, Tokuyama) scored mTBS data to ground enamel that were comparable to those measured for two one-bottle adhesives (One-Step, BISCO; Scotchbond 1, 3M) that involved a separate phosphoric-acid treatment. When the self-etch adhesives were directly bonded to unground, intact enamel, the resultant mTBS values, however, were significantly lower. Testing the marginal sealing potential and durability of the self-etching approach should obviously confirm these promising enamel performance data.
Clinical Bonding Effectiveness
At Leuven, the clinical effectiveness of adhesives has been routinely investigated in controlled tow-to-three-year follow-up studies using the same experimental protocol for almost 20 years. The retention rates shown in Figures 25a and 25b clearly illustrate the significant progress made in adhesive performance when adhesives (Figure 25a) from prior to versus after 1990 (Figure 25b) were used to restore cervical Class-V non-carious lesions with their respective restorative composite material. In part, this must be attributed to the introduction in the early '90s of the total-etch technique by which phosphoric acid is now also applied to dentin. Earlier adhesives often showed many failures within the first six months when applied strictly to dentin without any selective phosphoric acid etching of adjacent enamel (Van Meerbeek & others, 1994b). When following the same protocol in more recent clinical trials (total-etch systems were applied selectively to dentin), almost any early de-bonding failures were recorded. This must be attributed to a great extent to the enamel immediately adjacent to dentin always being (unintentionally) etched and guaranteeing a durable bond to the enamel margin. Adequate bonding to enamel alone may keep such restorations longer in place.
A great diversity in adhesives that can basically be categorized total-etch, self etch and glass ionomer adhesives exists. A clear trend exists towards simplified application procedures with a reduced number of application steps. However, simplification does not necessarily imply improved or even equal bonding effectiveness.
Conventional three-step total-etch adhesives remain the adhesives of choice for routine clinical use because of their least technique-sensitivity and their best laboratory and clinical effectiveness data. Today's major shortcomings are, amongst others, the relatively high technique-sensitivity of current systems and the apparent difficult-to-solve compromise to bond equally effective to enamel and dentin. Self-etch adhesives, either resin- or glass ionomer-based, may be most promising in overcoming these shortcomings. They do not require a rinse phase, which truly saves time and is less prone to manipulation errors. No discrepancy exists between demineralisation and infiltration. They offer a twofold bonding mechanism based on micro-mechanical interlocking through hybridisation to resist "acute" debonding stress and improved monomer-collagen interaction potentially by primary chemical bonding, which may be helpful to keep the bonds leakage-free in a long-term perspective.
An adhesive restoration, in conclusion, has many advantages over conventional non-adhesive restorative techniques except that it cannot yet be realized in a simple way.
Abate PF, Bertacchini SM, Polak MA & Macchi RL (1997) Adhesion of a compomer to dental structures Quintessence International 28(8) 509-512.