The use of glass polyalkenoate (GPA) cement in conjunction with composite resin using an acid etch technique is now well known and a 'sandwich' of tooth/GPA cement/composite resin has been recommended to form the cervical seal at the base of approximal boxes in Class II cavities. This paper reports on the 2-year results of a controlled clinical trial using a commercial GPA lining cement. The trial was designed to evaluate the efficacy of this 'sandwich' technique. Sixty-four restorations in the mouths of 22 patients were evaluated during the 2.5 year period of the trial. The composite resin component of the restorations performed well. Five restorations failed, all in the region of the approximal box. Four failures were related to the exposed GPA cement component of the 'sandwich'. Failure was probably related to placement difficulties of the technique. The use of GPA cement laminated with composite resin when the GPA cement was enclosed within the final restoration appeared to be a successful technique.

No abstract

The loss of material from specimens of three luting cements was measured after continuous erosion cycling in the laboratory. The glass ionomer luting cement showed significantly less material loss than the zinc polycarboxylate and zinc phosphate luting cements. Two hundred and fifty restorations cemented with one of the three materials were studied clinically for marginal integrity and retention over 3.5 years. The data were tested using survival analysis. Zinc phosphate cement gave the best clinical performance. Possible explanations for the poor correlation between the findings in the laboratory and clinical study are discussed.

Summary. The incidence of pulpal damage beneath cavities filled with ASPA, Chemfil and Chembond was studied in class I cavities in human premolar teeth after 1–10 days, 11–30 days, and 31 or more days. Cavities filled with Kalzinol or zinc oxide/eugenol acted as controls. All glass ionomer cements were accompanied by more pulpal damage than controls, but Chemfil was associated with the least changes. Bacteria were demonstrated at the material/cavity wall interface in a number of cavities with all materials, although no significant correlation was demonstrated between them and pulpal inflammation. The use of a base is recommended in cavities where large numbers of previously unaffected dentinal tubules are present.