Comparison of Full-arch Implant-supported Frameworks From an Intraoral Scanning or From a Conventional Impression

Eight consecutive patients ready for a full-arch rehabilitation on previously osseointegrated implants, and who had received an immediately loaded prosthesis, were selected. All implants had multiunit abutments placed on top. Cases from 4 to 6 implants were accepted, either upper or lower arches. Three sets of impressions were taken to each patient:

Conventional impression:

Open tray impression copings were screwed on top of all implants in the arch using a torque wrench set at 10 N/cm. Impression copings were splinted by means of Triad Gel clear resin, with an at least 3mm diameter in the resin connectors. Once the resin was photo-polymerized a cut was done in the centre of each connection with a 0.3 mm bur, to eliminate possible stress in the structure, and then it was splinted again with a drop of Triad Gel in each cut. A polyether impression was taken by means of a perforated plastic tray, and a master model was fabricated following the manufacturer's instructions. A plaster verification jig was fabricated on the model. After one week the patient came to the dental clinic where vertical dimension of occlusion was determined and the plaster key was screwed in the patient mouth to check passivity. In case the key had been fractured, a new impression would have been taken, and if not, final prosthesis could be processed.

Optical impression:

An impression of the antagonist and an impression of the edentulous arch with the healing abutments in place were taken by means of an optical scanner following the manufacturers protocol. The position of the implants was marked in the working standard tessellation language (STL) model, and the software cut a circle on top of those areas. The files were copied, and two datasets were obtained.

Scan-bodies were placed on top of each multiunit abutment, the first STL dataset was open and a second impression was taken, so that the scan-bodies were positioned on the working file.

Scan-bodies were then removed from the patient mouth, and temporary copings were placed on top of each multiunit abutment. A MedicalFit device was selected, and holes were drilled in the position of the implants until it fitted on top of all the cylinders. Cylinders were then splinted to MedicalFit device by means of Triad Gel clear. The device was then removed from the patient mouth and scan implant replicas were placed at the bottom of the metal cylinders. Finally, the device with the scan implant replicas was scanned with an intraoral optical scanner.

The following set of pictures were taken: Rest position, maximum smile (with the scan bodies placed in the mouth), front with retractors (with the scan bodies), and 45º with retractors (also with the scan bodies in place).

All pictures and STL datasets were sent to the dental technician to produce a temporary structure. The STL dataset from the scanning of Medicalfit device used was adjusted to the original STL dataset of the device in the library, and then superimposed to the dataset of the scan-bodies received from us, to allow repositioning of the scan-bodies and to correct possible discrepancies during the intraoral scanning process.

A temporary polymethylmethacrylate (PMMA) framework was delivered by the dental lab and tried in the patient mouth. Corrections were done to adjust occlusion, gingival contour and esthetic parameters where needed. Once esthetic parameters and occlusal scheme were considered correct, a new intraoral scanning was taken with the temporary framework in place. In case changes were done in the gingival contour, scanning of the gingival aspect of the PMMA framework was also done. The PMMA temporary was then left in the patient mouth.

Three pictures were taken with the PMMA framework placed in the mouth: Front full face at rest, front full-face smiling, front full face with separators and full face smiling at 45º.

Clinical adjustment assessment:

After one week two sets of the final prosthesis in full-zirconia were sent from the dental lab, one processed with conventional models and the second fully digital. Both sets were tested in the patient mouth by two calibrated blinded independent operators.

Misfit should be ideally measured in terms smaller than 10 microns, but clinical adjustment is difficult to assess using conventional or quantitative methods. In this research adjustment was assessed following four criteria: passivity in the insertion of the screws, tactile sensation, radiographs and screwing torque:

1. A Visual Analogue Scale (VAS) was used to assess the perception of passivity in the insertion of the prosthodontic screws. A ten centimeters line will be shown to the observer. One side of the line represents "perfect passivity", and the other "no passive at all". The observer will mark his passivity perception on the line between the two endpoints. The distance between "no passivity at all" and the mark will defines the passivity perception.
2. Examination of the marginal fit with an exploratory probe (#23/3 explorer) under 3,8x magnification. Three possible scores were possible: 0 (no gap perceived when probing), 1 (perception of the gap without entering it) and 2 (the tip of the explorer clearly entered the gap).
3. Periapical radiographs were taken with a positioning system to evaluate possible gaps. Five possible scores were possible from 1 to 5, being 1 no gap and increasing 0.15 mm each score until reaching 0.60 mm in score 5.
4. Digital photographs were taken with a 105 mm macro lens. The pictures were calibrated with the known height of the framework and the gap was measured. Four possible scores were possible, being 1 no gap and increasing 0.25 mm each score until reaching 0.75 mm in score 5.
5. Screwing torque on each abutment was measured with the app software from Ichiropro motor (BienAir, Bienne, Switzerland). All screws were hand screwed and then all but the most distal right were unscrewed. Torque was set at 15 N/cm and speed at 5 rpm. The most right distal screw was tightened first, and then the rest of the screws were tightened starting at the most mesial left, following to the most distal left, and from that to the next neighbor implant until all screws were tightened. The torque-time signature was studied in every screw to determine if the torque started to grow just at the end of the tightening or throughout the process. Three scores were possible: 1 for linear value with a fast increase at end of the tightening, 2 for soft continuous growth with a steeper increase at the end, 3 for steep increase at the beginning of tightening.

Once the assessment of the two frameworks was completed, the framework with better adjustment will be placed in the patient, after assessing the presence of correct occlusion, phonetics and esthetic parameters. The second set will be packed and delivered to the patient as a possible replacement in case of problems in the future.