Patients who received a metal framed ceramic screw retained implant crown or bridge were included in this study. These access holes were delimited only by ceramic. Those with metal surface exposure were excluded from this study. Patients with edentulous arch or section, full or partial denture as antagonists were excluded from the study. During the evaluation period, 2 patients (male) dropped out from this study. One access-hole (AMB/22) with an atypical shape (non-circular) causing a noticeable overfilling of the M4M was excluded from this study. Finally, 12 patients with a total of 56 access-holes were examined (28 access holes for both groups, CR and M4M). For each patient, access-holes were randomly divided in right and left sides to get equivalent numbers for both materials. All the fillings (CR and M4M groups) were performed by a single operator.

For each access-hole, the contact patterns of antagonistic cusp in the centric occlusion were recorded intra-orally. They included; A: contact in the center, B: contact on the border of the access-hole and C: no contact close to the border (Fig. 2). At the time of the restoration (T = 0), followed by 1, 3, 6, and 12 months (T = 1, 3, 6, and 12 M), impressions were taken with polysiloxane impression materials (President, Light Body, (lot: G01914), Soft Putty (lot: G08568), Coltène/Whaledent AG, Switzerland) in a double mixed method. Epoxy casts (Devcon ET, Lot: 350402, ITW PP&F, Japan) were made from these impressions, and the fillings were examined using a motorized digital microscope (DSX510, Olympus KeyMed Ltd, USA) with a 1 μm accuracy under ×30 digital magnification.

For each clinical recall (T = 0 to T = 12 M), intra-oral photographs of the access hole fillings were taken.

The surface areas of access holes could not be compared as the configuration of access-holes varied in each case. It was therefore necessary to compare longitudinal changes of the surface areas. At the time of T = 0, margin of the filling could not be clearly identified, thus, the measurement was fixed to begin at T = 1 M. For each sample, from T = 1 M to T = 12 M, the surface areas of access-hole filling were measured perpendicular to the vertical surface of the hole using the same digital microscope. The surface at T = 1 M was considered as 100%, and compared with the T = 3, 6, and 12 M of identical filling (Fig. 3a–e).