Root-analogue implants compared to forced orthodontic extrusion: a retrospective analysis of clinical, radiological and esthetic outcomes after restoration

For the present retrospective investigation, available clinical data regarding restorations supported by RAIs or natural roots after FOE were retrospectively evaluated and compared. Ethical approval was given by the local Ethical Committee of Charité – Universitätsmedizin Berlin, Germany (application numbers: EA4/140/18 and EA2/301/20). All RAI-treatments were performed by the author D.H. and respective follow-up examinations by the author M.W.H.B. All FOE-treatments and follow-up examinations were performed by the author M.B. To reduce subjective bias, analyzed outcomes were evaluated independently by both first and last specified authors of this article. All patients have received and signed a written informed consent form and patient information. It was conducted considering the STROBE statement for observational studies (https://​www.​strobe-statement.​org) where applicable.

For treatments with a RAI, the following criteria regarding the patients had to be fulfilled: 1) Non-smokers; 2) No medication affecting the bone metabolism; 3) Non-inflammatory surrounding soft and hard tissues and 4) Bone compartments surrounding the relevant tooth had to be intact. For FOE following inclusion criteria were defined: 1) Patients in need of a restoration of a deeply destroyed tooth with two adjacent teeth; 2) Probing depths ≤ 2 mm and defects violating the biologic width and/or a missing ferrule design preparation; 3) Prospective crown-to-root ratio ≤ 1; 4) Tooth mobility ≤ 1 and prospective single-tooth restorations. Teeth with suspicion of hypercementosis/ankylosis, molars and teeth with vertical root fractures were excluded. Treatments were only performed in patients whose compliance regarding the necessary monitoring was expected.

Due to the retrospective comparative study design, the cases to be evaluated had to be selected based on the available data. Therefore, n = 10 pre-molars restored with single crowns supported by RAIs, n = 10 severely damaged roots of pre-molars restored with single crowns after FOE, n = 10 anterior teeth (canine to canine) restored with single crowns supported by RAIs and n = 10 severely damaged roots of anterior teeth restored with single crowns after FOE were selected. Thus, 40 cases could be included. Available data were anonymized, retrospectively evaluated and statistically analyzed. Post-hoc power-analysis was performed with a free-to use software to calculate statistical power (G* Power 3.1.9.7, Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany) with a sample size of n = 20 per group, α = 0.05 and an effect size of 1.17 resulting in a power of 0.97 (97%).

For rehabilitations with RAIs, these had to be manufactured individually before surgeries. Impressions were taken with customized trays and a polyether material (Impregum, 3 M Deutschland GmbH, Neuss, Germany). These were cast with type IV plaster and digitized with a laboratory scanner where after the model data were available as Standard Triangle/Tessellation Language (STL) datasets. Bite records were taken with an injectable elastomeric A-silicone (Futar D, KettenbachGmbH&Co. KG, Eschenburg, Germany) and DICOM data were exported from an x-ray device (PAX i-3D, VATECH, Hwaseong-si, Gyeonggi-do, South Korea). A highly trained specialist of the manufacturer superimposed the DICOM and STL data and designed the Temporary Protective Covers (TPCs), implant and abutment portions of the RAIs (Fig. 1e and f) and Try-Ins (exact copies of the RAIs made from sterilizable resin, Fig. 1c and d). From resulting Computer Aided Design (CAD) data, TPCs and RAI components were milled in a Computer Aided Manufacturing (CAM) process. All ceramic parts (abutment portion, root portion in case of all-ceramic RAIs and TPCs) were made of yttria-stabilized tetragonal polycrystalline zirconia (Y-TZP) and sintered. With hybrid RAIs, root portion was manufactured from pure titanium (medical grade IV) and bonded to the aforementioned abutment portions made from Y-TZP in a special oven process using a biocompatible glass material without any voids. These connections were checked using x-rays within quality control. Afterwards, root portions of both (all-ceramic and hybrid RAIs) were modified with macro- and micro retentions (medical grade corundum and etched). The Benex Extraction-System (Benex Extraction-System, Helmut Zepf Medizintechnik GmbH, Seitlingen-Oberflacht, Germany, Fig. 1b) was used to remove teeth minimally invasive in vertical direction. If this was not successful, teeth were carefully removed using conventional extraction techniques. Intact surrounding bone compartments were obligatory. The expected fit of RAIs was tested by using the respective Try-Ins (Fig. 1d). Before implant insertion, implant surfaces and alveoli were wetted with Plasma Rich Growth Factors (PRGFs; BTI Biotechnology Institute, San Antonio, Spain). RAIs were carefully placed (Fig. 1e) and present voids were filled with PRGFs as well. As load protection TPCs were semi-adhesively attached (RelyX Ultimate, 3 M Deutschland GmbH) to one or both adjacent teeth depending on the design. This provided a protective gap of approx. 0.6 mm between the abutment portion of RAIs and TPCs it selves. The workflow has already been described and illustrated in detail in a publication in 2020 [28] and is additionally shown in Fig. 1. Healing was usually checked 3–6 month after surgeries using clinical and radiological parameters. Prosthodontic treatments were performed in the office of the author D.H. and different referring dentists, whereby no constantly defined workflow was followed. In total, 16 hybrid and four all-ceramic RAIs with respective restorations were investigated.

Prior to orthodontic extrusion, probing depths, tooth mobility, defect size and radiographic images were assessed to determine the amount of extrusion and the prospective crown-to-root ratio for each patient individually. Interocclusal available space for orthodontic extrusion was analyzed with the aid of gypsum models. After removal of insufficient restorations, a fiber-reinforced composite-based post (Extrusion pin, Komet Dental, Lemgo, Germany) was placed on the root surface of respective teeth with self-adhesive resin (RelyX Unicem 2 Automix, 3 M Deutschland GmbH, Fig. 2c) in central vestibulo-oral direction, at the widest root diameter.

A second post serving as anchorage was adhesively bonded to adjacent teeth with flowable composite resin (Tetric EvoFlow, Ivoclar Vivadent AG, Schaan, Liechtenstein, Fig. 2c). Elastics were placed in the orthodontic appliance to initiate forced orthodontic extrusion with forces > 0.5N (Fig. 2c). At the same appointment, supra-crestal fibrectomy, scaling and root planning procedures were performed [54]. In some clinical cases two bars had to be bonded to adjacent teeth due to minor occlusal space. In clinical situations requiring crown restorations on neighboring teeth, anchorage was realized with the aid of a provisional FDP and a reduction of the pontic area to allow for a sufficient amount of extrusion. Clinical details on the workflow have already been described and published [47]. An exemplary workflow is additionally shown in Fig. 2. Elastics were changed by the patients twice daily due to the loss of tension. Patients were available for control visits once a week. The mean amount of extrusion was 3.50 ± 0.87 mm and the mean time of extrusion 17.88 ± 10.98 days. After extrusion, teeth were bonded to adjacent teeth with composite resin (Tetric EvoFlow, Ivoclar Vivadent AG, Fig. 2d) for a retention period of at least 8 weeks (mean time of retention 130.60 ± 89.12 days) to prevent orthodontic relapse [46]. After revision of endodontic fillings, placements of glass-fiber posts (X-Post, DentsplySirona, Bensheim, Germany) were performed where indicated, teeth were built up with composite resin and restored with single crowns.

Within the framework of a feasibility analysis regarding this investigation, existing data sets were reviewed for completeness by the authors M.B. and M.W.H.B. to be able to compare restorations supported by RAIs or natural roots after FOE. The novel Functional Implant Prosthodontic Score (FIPS) was chosen [55] for evaluation and slightly modified regarding rehabilitations of natural roots after FOE not representing implant-retained restorations (Table 1). It was selected as it is a straightforward, self-explaining, reliable, reproducible and quickly applicable score [55, 56].

FIPS is defined by five variables, allowing to evaluate the interproximal area, the occlusion, the design, the mucosa and the bone. After rating every mentioned parameter with 0, 1 or 2, a maximum score of 10 can be achieved by a single restoration (Table 1). In case of subcategories (i.e. “interproximal” with “contacts” and “papillae”) with different ratings, the lowest has to be selected. Therefore, for evaluation, following data had to be present: 1) Photographs for the visual assessment of papillae, the design/esthetics of the restorations, and the mucosa; 2) Documented information regarding the interproximal contacts (checked with dental floss); 3) Documented information regarding the static and dynamic occlusion (checked with shimstock foil, thickness: 8 μm); 4) Available x-rays for the evaluation of marginal bone loss.

For evaluation of interproximal bone changes regarding restorations of severely damaged roots after FOE, radiographic images of teeth before intervention and at recall appointments were superimposed. Therefore, reference lines were drawn on apices and incisal edges of neighboring teeth. The root length of the extruded tooth was divided into ten equal segments. An interproximal bone change of one tenth in a section was defined as interproximal bone loss [57]. A stable interproximal bone level was rated with 2, a change in one tenth of a section was rated with 1 and an interproximal bone loss of more than one tenth of a section was rated with 0 (Fig. 3).

Due to the availability of the implant lengths of RAIs, the evaluation of bone loss could be carried out in accordance with FIPS score after superimposition of pre- and post-loading x-rays using mathematical rule of three (Fig. 4). Total marginal bone loss without a view to the rough-smooth border between implant and abutment portions were rated. Due to the retrospective study design, no individual bite holders could be used for both investigated concepts.

All statistical tests were performed by an independent examiner using “SciPy” (https://​scipy.​org/​, last accessed 13th of March, 2022), a Python-based open source software environment mainly used for scientific analysis, visualizations and related activities. Krippendorff’s alpha coefficient was calculated to assess inter-rater reliability between the authors M.W.H.B. and M.B. Due to ordinal FIPS data, Mann–Whitney-U-Test was used to compare the assessed parameters regarding restorations of natural roots after FOE or supported by RAIs. Level of significance was set to p < 0.05.