Introduction
Screw-retained restorations, whether single units or hybrid prostheses, have become widely used when restoring implants. These have several benets compared with cemented restorations. Cemented restorations have the potential to leave residual cement subgingivally in the sulcus, thereby possibly causing inammation and resulting in bone loss around the implant crestally (peri-implantitis). Additionally, with a cement-retained restoration, the screw attaching the abutment to the implant may loosen, requiring access to be created through the restoration or destruction of the restoration to access the screw and tighten it. Screw-retained restorations circumvent the potential subgingival inammation related to residual cement by eliminating intra-oral cementation of the restoration to the abutment. Also, should the screw loosen or the restoration require removal for periodontal maintenance or another reason, an access hole at the occlusal or lingual portion of the restoration permits access to the xation screw. The composite, as well as cotton pellet or PTFE tape, sealing the access hole is removable,
and the screw can be tightened or removed.

Traditionally, for screw-retained restorations, a titanium-base abutment (Ti-base) is placed on a laboratory analogue on the model at the laboratory. The restoration is then fabricated to seat over the Ti-base and is luted to the Ti-base with a resin cement (Fig. 1). This is sent to the practitioner, who inserts it, tightens the xation screw (Fig. 2) and then seals the access hole with either a cotton pellet or PTFE tape and then composite. Problems are unfortunately found with Ti-base restorations. The restoration may separate under function from the Ti-base it was luted to. One cause of this is insufcient height of available Ti-bases from the implant manufacturers, so less surface area is present where it is luted to the overlying ceramic restoration. When the prosthesis presents with separation between the Ti-base and ceramic of one or more units, the treatment is to re-lute the crown to the Ti-base intra-orally. This can be a challenge because the crown may not fully seat on to the Ti-base, increasing occlusal contact on the implant restoration, which may require occlusal adjustment of the restoration or, if minor and not felt by the patient, may result in increased loading on the implant and subsequent bone loss over time. Additionally, if left intra-orally, residual cement can lead to periodontal issues as reported with cementable implant restorations. To avoid this, once the crown has been recemented on the Ti-base intra-orally and the cement has set, the restoration is removed, the cement cleaned from the junction of the Ti-base and ceramic, and the restoration then reinserted intra-orally. This is a time-consuming procedure chairside. With regard to a hybrid prosthesis, the entire zirconia portion of the prosthesis may separate from all of the Ti-bases intra-orally, necessitating re-luting on to the Ti-bases intra-orally. In some clinical situations, one or more of the Ti-bases may separate while the others remain connected intra-orally (Fig. 3).

This may not be detectable by the patient and will essentially decrease the number of implants supporting the prosthesis, and the units that have had Ti-base separation will allow micro-movement in the vertical access. This may lead to other units undergoing Ti-base separation. Laboratories have attempted to eliminate the use of Ti-bases, having the zirconia contact directly to the multi-unit abutment (MUA) with no metallic component in the prosthesis at the connection (Fig. 4). The xation screws for the implants on the market have a at bottom designed to seat in the Ti-base. This causes stress concentration at the bottom of the screw head, where the zirconia is thin (Fig. 5). As stress continues, fracture may result at the thin portion lateral to the MUA or restoration margin, leading to restoration failure (Figs. 6 & 7). These screws also do not create tensile load when placed into the implant with or without a Ti-base present, leading to potential screw loosening, even when torqued to the manufacturer’s specications with a torque wrench.

The Rosen screw
To overcome those inherent problems associated with Ti-bases (separation of the Ti-base from the restoration), screw loosening and elimination of the Ti-base with use of conventional screws and resulting restoration marginal fracture, the Rosen screw (www.rosenscrew.com) was developed. The Rosen screw eliminates the use of Ti-bases so that nothing is luted within the restoration that may separate under function over time. Depending on design, this screw can connect the screw-retained restoration to MUAs on the implants or directly to the implants with no intermediate MUAs present. This increases the thickness of the zirconia, decreasing its fracture potential under loading. The Rosen screw has a tapered head, eliminating the at bottom of conventional screws (Fig. 8). When the screw is threaded into the implant through the restoration, lateral tension locks the restoration to the implant, providing tensile load and eliminating potential screw loosening under functional loading (Fig. 9). The geometry of the screw head locks the prosthesis and can be achieved with hand tightening alone without use of a torque wrench. The screw can be used for prostheses fabricated from resin, PMMA, graphite or zirconia. Screw cost is comparable with other screws offered by the various implant manufacturers. Rosen screws are available in three sizes, colour-coded to make identication visually easy—1.40mm (blue), 1.60mm
(gold) and 1.72mm (silver)—and are compatible with most implants available on the market (Figs. 10–12).

Libraries are available for design of the restorations in-ofce or in-laboratory for milling to accommodate use of the Rosen screw (https://rosenscrew.com/libraries/). These include exocad, Imetric4D, Pic Dental, Zirkonzahn and 3Shape.
Restorative examples
With regard to hybrid restorations, the prosthesis is designed to fit to the MUAs, providing thicker material over the MUAs to resist fracture potential under loading (Fig. 13). The occlusal view of the hybrid prosthesis looks similar to that for which Ti-bases have been used, having a small access hole over each implant on the prosthesis (Fig. 14).

Radiographically, the design demonstrates an intimate t to the MUAs and increased prosthetic material surrounding the MUAs for improved fracture resistance of the prosthesis (Figs. 15 & 16).

Case example
A 54-year-old male patient presented desiring implants to replace failing dentition in both arches. Examination and radiographs conrmed that the remaining dentition was in poor condition (Fig. 17). Treatment was discussed with the patient and would include extraction of the remaining dentition, placement of six implants in the maxilla and six implants in the mandible, and restoration with screw-retained hybrid prostheses in both arches, utilising a digital workow with Imetric (Imetric4D). The patient returned for surgical treatment of the maxillary arch, which consisted of extraction of the remaining dentition under local anaesthetic and placement of Tuff implants (Noris Medical Inc) into six sites. MUAs were placed on the implants and the soft tissue closed with sutures. Impression copings were placed on the MUAs and an impression taken of the arch using a Rosen tray that articulated with the teeth in the mandibular arch. A provisional maxillary hybrid prosthesis was fabricated without Ti-bases and attached intra-orally with Rosen screws. For the mandibular arch, the patient was treated at another appointment, during which the extractions were done and six Tuff implants were placed. The implant placed in the site of the mandibular left second premolar had insufcient insertion torque so could not be immediately loaded with the provisional hybrid prosthesis. Like with the maxillary arch, MUAs were placed on the ve implants that would be loaded and an impression taken with a Rosen tray to fabricate an immediate provisional mandibular hybrid prosthesis. The provisional prosthesis was fabricated without Ti-bases and attached intra-orally with Rosen screws. The patient returned after three months of integration for nalisation of the prostheses. During that time, the patient had suffered a minor stroke, and it was thus decided not to subject him to surgery to uncover the buried implant, particularly since the mandibular prosthesis was functioning well on the ve implants that had been loaded. The provisional hybrid prostheses were removed, and an impression was taken of the implants and healed soft tissue for fabrication of the nal zirconia hybrid restorations. The provisional restorations were reinserted, and the impressions sent to the laboratory for fabrication of the nal prostheses. The zirconia hybrid prostheses without Ti-bases were returned from the laboratory, and the patient returned for insertion (Figs. 18 & 19).

The provisional prostheses were removed (Fig. 20) and the zirconia hybrid prostheses inserted intra-orally and fixated with 1.4mm Rosen screws that were tightened by hand. The screw access holes were then lled with a piece of PTFE tape and sealed with a owable composite resin and light-polymerised. The occlusion was checked and adjusted as needed, completing the treatment (Fig. 21).

Conclusion
Screw-retained restorations may present with issues related to marginal fracture of the prosthetic material, separation of Ti-bases from the overlying restorative material and screw loosening. The Rosen screw eliminates those issues with its tapered screw head design and elimination of the use of a Ti-base. The screw may be tightened by hand without the need for a torque wrench, and sufcient tensile load is present to prevent screw loosening. Rosen screws are designed for use with splinted implants either in a bridge or full-arch hybrid prosthesis. The screws cannot be used in single-unit screw-retained crowns, as the MUA offers no anti-rotational mechanism between the crown and MUA.