Prosthesis-Based 3D Guide System for Maxillary Implant Placement: A Feasibility Study Using a Split-Mouth Evaluation

Background/Objectives: To evaluate the clinical feasibility and short-term radiographic outcomes of adapting a prosthesis-based guide system originally developed for single-implant placement for multi-implant placement in the edentulous maxilla, and to compare implant-level marginal bone change from prosthetic loading to one year between full-rough (IBT/IBNT/IBR) and hybrid-surface (MSC-IBT/MSC-IBNT) implants in a split-mouth design. Patient-reported outcomes were assessed with the OHIP-14 questionnaire. Methods: Fifteen patients with an edentulous maxilla received four or five maxillary implants placed flaplessly using a 3D-printed prosthesis-based guide incorporating polyether-ether-ketone (PEEK) rails and interchangeable angulation-correction sleeves (0°, 12°, 24°). Implants had either a fully roughened or a hybrid (rough apical and middle third with a machined coronal collar) surface. Side allocation was non-randomized: the hybrid side was assigned by alternating sequence and three patients received only full-rough implants for prosthetic reasons. All patients followed a delayed loading protocol and received a screw-retained zirconia fixed bridge. Standardized periapical radiographs were obtained at definitive prosthetic loading (baseline) and at the 12-month follow-up. Implant-level marginal bone loss was calculated as the mean of mesial and distal measurements per implant. All radiographic measurements were performed by two independent examiners (M.T. and T.M.); intra-rater reliability (M.T., remeasurement of 10 radiographs) and inter-rater reliability (M.T. versus T.M., full dataset) were quantified by intraclass correlation coefficient (ICC; two-way mixed effects, absolute-agreement, single-measurement). The primary split-mouth surface comparison was performed at the patient level by averaging implant-level change scores within each surface for each patient and comparing the paired patient-level means with a Wilcoxon signed-rank test. No a priori sample-size calculation was performed; the study was designed as a feasibility pilot. Results: Sixty-one implants were placed in 15 patients (seven male, eight female; mean age 62.5 ± 8.9 years; three current smokers). Four implants in three patients required removal and replacement during the observation period (three early failures, one late failure; per-implant early-failure rate of 3/61, 4.9%); one patient (P7) withdrew from clinical follow-up. Paired baseline and 12-month radiographs were available for 53 implants from 14 patients. Median implant-level marginal bone level increased from 0.38 mm (IQR 0.20-0.54) at baseline to 0.78 mm (IQR 0.47-1.32) at 12 months (paired Wilcoxon signed-rank, p < 0.001); two implants exceeded 4 mm of bone change at 12 months. In the patient-level paired surface comparison (n = 8 patients contributing at least one full-rough and one hybrid implant with paired data), full-rough implants showed less 12-month marginal bone change than hybrid implants in every paired patient (median paired difference full-rough hybrid of -0.49 mm; Hodges-Lehmann pseudo-median of 0.55 mm; paired Wilcoxon p = 0.012). OHIP-14 scores at one year (n = 14) showed a pronounced floor effect, with most patients scoring zero across most domains. Both intra-rater (M.T.) and inter-rater (M.T. versus T.M.) reliability showed good agreement (ICC = 0.85). Conclusions: A prosthesis-based guide system originally validated for single-implant placement can be feasibly adapted for flapless multi-implant rehabilitation of the edentulous maxilla, with early clinical and radiographic outcomes broadly consistent with comparable published series. Contrary to the design rationale that a machined coronal collar would limit early crestal remodeling, full-rough implants showed less 12-month within-patient bone change than hybrid implants in the eight paired patients; this finding is preliminary and hypothesis generating given the small, unbalanced paired sample and the contrast with larger published series. The approach is best characterized as a reduced infrastructure alternative to proprietary guided-surgery platforms, remains operator dependent, and requires confirmation in formally powered, balanced split-mouth trials with concealed allocation, placement anchored bone level measurement, postoperative CBCT for deviation quantification, and longer follow-up.