Double Crown-Retained Removable Prostheses Supported by Implants or Teeth and Implants: A Long-Term Clinical Retrospective Evaluation

• Double crown telescopic dentures using implants or teeth-implants as abutments are a good option for the restoration of partially or completely edentulous arches, avoiding extensive bone augmentative surgeries.

• Double crown telescopic dentures with distal cantilevers produce better results in terms of technical complications compared to those with distal extension saddle.

The rehabilitation of completely edentulous (CE) or partially edentulous (PE) patients with implant (I)-retained prostheses or tooth-implant (TI)-retained prostheses can be challenging [1‒4]. The feasibility of available restorative options depends on anatomical considerations, including usable inter-arch space, which inform strategic decisions to maintain or extract teeth for implant placement [5‒7]. Patient-dependent factors, such as individual preferences, expectations, and financial/insurance considerations, also influence treatment planning.

One treatment option is to use removable dental prostheses (RDPs) based on double crowns (DCs, telescopic), achieved with splinting implants, teeth, or teeth and implants [1, 3, 5, 7]. PE jaws with few remaining abutment teeth can be restored optimally with tooth-implant-retained double crown removable dental prostheses (TI-DC-RDPs) [5‒9]. In CE arches, implant-retained double crown removable dental prostheses (I-DC-RDPs) are an effective attachment-retained overdenture option [10‒13]. DCs were first placed on natural teeth in the 1960s and have been widely applied for denture support [1, 3, 7, 14]. This approach allows easy access for oral hygiene maintenance [4, 9, 12, 15, 16], and in the event of abutment (natural tooth or implant) loss, the system may be adapted to achieve sufficient support for RDPs [3, 14, 17, 18]. Telescopic attachments provide excellent retention owing to the frictional interaction between the primary crowns (PrCs) and secondary crowns (SecCrs) [5, 13, 17]. Hard- and soft-tissue deficits can be managed with acrylic and/or ceramic/composite veneer materials, depending on the prosthesis design [1, 3, 5, 17, 19]. However, limitations for these restorations include inter-arch space requirements for construction and extensive tooth preparation, as well as meticulous, technique-sensitive fabrication processes.

Types of DCs include rigid parallel-walled telescopic crowns, rigid tapered conical crowns, rigid tapered electroformed gold crowns, and nonrigid (or resilient) telescopic crowns with clearance fit. Additional variations in construction method, materials, and retentive features are available [5, 19‒24]. Several investigators have assessed factors affecting the survival of teeth and implants with I- and TI-retained RDPs, including DC system type, numbers, and positions of abutment teeth and implants [1, 5, 9, 14, 24‒27]. Most studies of outcomes for I- and TI-RDPs with rigid telescopic designs have been retrospective and conducted on 20–30 patients; larger group and long-term follow-up (>10 years) data are scarce [3, 5]. Various RDP framework materials and fabrication techniques have been used, and the clinical performance in terms of biological and mechanical complications and maintenance requirements has been compared [3, 4, 10, 12, 19, 28, 29]. To the best of our knowledge, there are no reports comparing complication types and rates between metal-composite resin cantilever (CAN) and hybrid metal-composite resin/acrylic resin free-end saddle (SAD) design-based DC-RDPs.

The primary objective of this retrospective study was to investigate the long-term survival of implants and natural tooth abutments with I- and TI-RDPs using rigid electroformed gold DC designs. Technical and biological complication types and frequencies were compared between TI-DC-RDPs and I-DC-RDPs. A secondary objective was to evaluate and compare prosthetic complications and periodontal/peri-implant outcomes between CAN and SAD design DC-RDPs, for TI-DC-RDPs and I-DC-RDPs. The null hypotheses were that there would be no differences in terms of biological and technical complication rates between (i) TI-DC-RDPs and I-DC-RDPs and (ii) CAN and SAD designs for both TI- and I-DC-RDPs.

In this private practice-based, non-randomized retrospective study, data from 53 PE and 57 CE arches in 110 nonsmoking patients (59 males and 51 females; mean age 53.9 years [range: 35–76]) restored with DC-RDPs during the period 2000–2016 were analyzed. The sample included I- and TI-DC-RDPs, supported by a total of 508 dental implant abutments and 153 natural tooth abutments (Table 1). All patients underwent dental extraction and periodontal treatment at least 3 months prior to the commencement of the treatment presented in this report.

Patients were informed orally and in writing (i.e., informed consent) about the planned treatment procedures and given at least 2 weeks for consideration. Patients had the right to withdraw consent and to interrupt treatment at any time without repercussions. For each patient, the treatment plan and the informed consent were approved by the national health authorities, which also approved the analysis of the cases and the publication of the results (Dental Council North-Rhine; Germany; No.: RA 232.20 AK-cls). All procedures were performed by the first author.

Selection criteria were as follows: (i) patients who requested RDPs without maxillary palatal coverage or a mandibular lingual bar/plate; (ii) no contraindication for implant surgery; (iii) no epilepsy, uncontrolled hand twitching, or hand tremor; (iv) no bruxism and/or jaw clenching; (v) adequate oral hygiene (plaque less than 10% and bleeding on probing less than 8%) at the beginning of implant surgery and/or restoration; (vi) endodontically treated teeth were not used as the most distal abutments; (vii) only one restored arch per patient was selected randomly for inclusion; (viii) second or 3rd molars were extracted, no implant was placed in these areas, and all restorations ended at the first molar position; (ix) patients who were unable to attend the recall appointments or not willing to participate in the study; and (x) patients with incomplete electronic documentation of the implant placement, prosthesis fabrication process, and/or follow-up appointment information.

Initially, 117 patients were enrolled for evaluation in the study. Applying the abovementioned inclusion and exclusion criteria, 7 patients were excluded from the list. Two patients did not consent to participate or were unable to attend the recall appointments, 2 patients could not meet the set oral hygiene requirements, and 3 had distal root canal treated teeth. Thus, 110 patients were restored definitively in one or both arches with I- or TI-DC-RDPs. All restored patients (n = 110) in the study had a minimum follow-up period of at least 3 years with a maximum of 16 years.

Implants were placed in accordance with prosthetically driven protocols: either one-stage (using Straumann Standard Plus [solid screw], 10 mm of length, Straumann Inst., Basel, Switzerland) or two-stage (using Camlog root line, 11 mm of length; Camlog Biotechnologies, Switzerland). They were uncovered and loaded 5 months after placement. The study sample (n = 110) included 2 main groups: group A (CE) (n = 57) and group B (PE) (n = 53), involving maxillary (mxl) or mandibular (mdl) DC-RDPs. The prostheses were provided either with CANs (unilateral or bilateral) or SADs (unilateral or bilateral) and subdivided into subgroups (A1, A2, A3, A4, B1, B2, and B3) based on the arch type (mxl; mdl) and design (CAN and SAD) as shown in Tables 1 and 2.

When the 2nd premolar position could serve as the site for the last abutment tooth or implant, the 1st molar served as a single CAN tooth for CAN DC-RDP placement (unilaterally or bilaterally). The CAN design involved the Co-Cr cast framework with extension supporting composite resin pontic teeth without flanges. When bone support in the 2nd premolar/1st molar region was deemed to be inadequate and the patient declined augmentative surgery, implants were placed up to the 1st premolar position and the two posterior teeth were replaced (unilaterally or bilaterally). In this case, the distal framework extension parts were designed as cast meshwork to support an acrylic base with prefabricated denture teeth in the free-end SAD area resting on the residual foundational tissue of the alveolar ridge.

Impressions were recorded using an open-tray technique with polyether impression material (Impregum; 3M ESPE, St. Paul, MN, USA). Natural teeth received cast-gold PrCs, and implants received system-specific UCLA customized implant abutments (CIAs). PrCs and CIAs were fabricated with a convergence angle of 2° and cast using a gold alloy (Portadur P4, 68.50% gold; Wieland Dental, Pforzheim, Germany) and served as primary telescopes (PrTEs). Electroformed 0.25-mm-thick gold copings (Auro Galvano Crowns [AGCs], Galvano gold, >99.9% gold; Wieland Dental) were fabricated for all abutments and served as SecCrs. Framework castings were constructed in Co-Cr-Mo alloy (Ankatit; Ankatit-Anka Guss, Waldaschaff, Germany).

At try-in, PrTEs and SecCrs were transferred and positioned on abutments. Consequently, frameworks were placed over the SecCrs and verified with occlusal records. The jaw relationship was re-recorded with a central tracing device and facebow mounted on a semi-adjustable articulator (SAM 2P; SAM Praezisionstechnik, Gauting, Germany) and occlusion adjusted as required. Frameworks were veneered with a micro-ceramic composite (Ceramage; Shofu, Ratingen, Germany). Veneering of free-end SADs was made using an autopolymerizing acrylic resin base material (PalaXpress Ultra; Heraeus-Kulzer, Hanau, Germany) and prefabricated acrylic teeth (SR; Ivoclar Vivadent, Ellwangen, Germany).

Finally, restorations were tried-in before finishing and polishing. The PrCs were luted to abutment teeth with zinc phosphate cement (Harvard; Harvard Dental, Hoppegarten, Germany), and the CIAs were torqued to 35 Ncm. SecCrs were placed on the PrTEs, and the supra-construction was luted with a self-curing compomer cement (AGC Cem; Wieland Dental) and allowed to set for 3–4 min. Occlusion was re-evaluated and restorations were finished, polished, and delivered.

Gingivitis and mucositis were treated with debridement and oral hygiene improvement. Lost teeth and implant abutments were replaced with implants only when no distal adjacent abutments were present. Periodontal disease was classified according to Armitage [21]. Periodontal pockets with probing pocket depths (PPDs) ≤5 mm and/or attachment loss (AL) ≤3 mm (periodontitis 1) were treated with scaling/root planning. Otherwise, periodontal surgery was performed (periodontitis 2). Peri-implant mucositis and peri-implantitis were diagnosed according to Mombelli and Lang [15] and Lindhe and Meyle [22]. Areas with vertical bone loss affecting up to 50% of the implant length (peri-implantitis 1) were treated with augmentative surgery. When vertical bone loss >50% of the implant length (peri-implantitis 2) was observed, implant removal and subsequent augmentation of the bone defect were undertaken. Artificial tooth loss, veneer fractures, and free-end SAD fractures were repaired in the dental laboratory. In the cases of implant failure, the DC-RDPs continued to function after the SecCrs were filled with resin. In the cases of tooth loss, the teeth were replaced with implants.

After DC-RDP delivery, PE patients were enrolled in supportive periodontal care (SPC), comprising quarterly annual follow-up appointments, wherein restorations were polished and oral hygiene was assessed and reinforced. Additionally, patients’ abutments and DC-RDPs were examined for technical or biological complications.

Comparisons were made between groups A and B and among subgroups, with analyses performed at the patient level. Continuous variables were assessed for normality using the Shapiro-Wilk test. Categorical variables are presented as numbers with percentages and compared using the Pearson’s χ² test. Outcome variable data were included in the survival analyses over the follow-up period. Hazard ratios (HRs) were calculated to compare complication and recession risks between groups, among subgroups, and according to distal extension and implant types. The assumption of proportional hazards was validated using a log-log plot. Time intervals from loading to each outcome, capped at the previous follow-up, were calculated. Kaplan-Meier analysis was used to establish the cumulative proportions of groups for which the target outcome had occurred at various follow-up timepoints, using the patient as the unit of analysis. Cox regression was used to compare outcomes between patient subgroups and to examine the association of complications with recession (from a baseline assumption of 0). Statistical analyses of the data were conducted using Stata software (v. 15.1; StataCorp LLC, College Station, TX, USA). The sample size of this study was determined based on previous similar studies [9, 30] comparing I-DC-RDPs with TI-DC-RDPs. A post hoc power analysis based on the differences between the abutment (implants and teeth) and prostheses survival rates, in terms of the number of complications observed with the two main modalities (I-DC-RDPs and TI-DC-RDPs), confirmed 90% power for the study [31].

The distribution of DC-RDPs by group-subgroup and patient-related demographic characteristics are shown in Table 1. The distribution of 110 DC-RDPs according to the arch and distal extension restoration design is presented in Table 2. In total, 153 natural teeth and 508 implants (421 Straumann and 87 Camlog) were used as abutments (Table 2). Two out of 153 teeth were lost, yielding a 10-year tooth survival rate of 99.3% (95% CI: 95.4–99.9%), and 3 of 508 implants (peri-implantitis 2, [2 Straumann and 1 Camlog]) failed, yielding a 10-year implant survival rate of 99.3% (95% CI: 97.5–99.7%). Ten-year survival rates for Straumann and Camlog (n = 9) implants were 99.3% and 98.8%, respectively. The estimated rates of implant failure and natural tooth loss at 15 years were 3% (95% CI: 1–9%) and 2% (95% CI: 0–7%), respectively (Fig. 1).

Frequencies of technical and biological complications and recession rates are reported in Table 3. Notably, no framework or abutment fractures, abutment screw loosening, or tooth intrusions were observed. Overall, nearly one-fifth of patients experienced technical complications, while approximately a third of patients experienced biological complications. Periodontitis and gingivitis occurred more commonly in group B than in group A (Table 3). Conversely, mucositis and peri-implantitis occurred more commonly in group A than in group B (Table 3). Recessions were observed in 87% of restorations (Table 3). Recessions were more common than technical or biological complications, being observed in nearly half of the arches within each group by the 5-year follow-up and affecting nine-tenths of all patients at 10 years (Table 3). The median durations of recession occurrence were similar in groups A and B. In the total sample, recession was observed in more restorations than technical or biological complications.

Generally, complications occurred within 10 years. Technical and biological complications were more frequent with TI-RDPs than with I-RDPs. The percentage of technical complications increased from 5 to 15 years (in group A: 4%, 14%, and 17% and in group B: 13%, 23%, and 23%; at 5, 10, and 15 years, respectively; Table 3). The percentage of biological complications also increased (group A: 14%, 31%, and 31%; group B: 23%, 38%, and 38%, at 5, 10, and 15 years, respectively; Table 3). The cumulative rates of DC-RDPs free of technical complications were 86% (CI 95%: 73–93%) for I-RDPs and 77% (CI 95%: 63–86%) for TI-RDPs. There were no significant differences (p > 0.05) in technical and biological complications between the two main groups A (I-DC-RDPs) and B (TI-DC-RDPs) (Table 4). The differences were also insignificant between the different cantilevered and saddled prosthesis designs for I- and TI-DC-RDPs (p > 0.05), except for the technical complications between I- and TI-RDPs, where there were significant differences noted (p < 0.05) (Table 4). The technical complications were 4 times more frequent in the mxl group B SAD restorations (B2) compared with the mxl group B CAN (B1) or mdl CAN (B3) restorations (Table 4). Overall, the risk of developing a technical complication over the follow-up period was similar between the A and B groups and across implant types and prosthetic designs (Table 4; Fig. 1). The risk of technical complications could not be compared between subgroups defined by distal extension type due to the lack of such complications in one group (Table 4). Table 4 also illustrates that biological complication risk values did not differ across groups or subgroups, distinguished by distal extension or implant type. The Kaplan-Meier plots (Fig. 1) further depict the time-related differences in complications for both I- and TI-DC-RDPs.

The risk of gingival recession occurrence was similar in groups A and B (Fig. 1) and for the different subgroups in group B. The differences were also not significant between different distal end variations and implant types. The frequency of gingival recession was about 0.57 times in groups A2 and A4 compared with groups A1 and A3 (p = 0.05; Table 4).

In this private practice-based, non-randomized retrospective study, clinical data from 110 nonsmoking patients receiving DC-RDPs based on I or TI abutments for a period of up to 16 years were assessed to determine abutment and prosthesis survival rates and complication types and rates. No significant differences were found in terms of technical and biological complications between the two main treatment modalities, I-DC-RDPs and TI-DC-RDPs, thus affirming the first null hypothesis of the study. There were significant differences found in technical complication rates between the CAN and SAD designs for the TI-DC-RDPs, hence rejecting part of the second null hypothesis. However, the differences in complication rates between the CAN and SAD I-DC-RDPs and the biological complication rates of TI-DC-RDPs were not significant, hence the other part of the second null hypothesis failed to get rejected.

The primary reasons for DC prosthesis selection in this study included anatomical and/or financial considerations. Two prosthesis designs were tested in this study with regards to the use of different combinations of materials and tissue coverage. The CAN design had the composite resin bonded to the cast Co-Cr framework, with CAN framework extensions supporting composite resin pontic teeth without flanges. In the SAD design, the framework extensions in the distal extension parts were designed as denture retentive elements continuous with the metal-composite resin RDPs to support an acrylic denture base with prefabricated denture teeth in the free-end SAD area. The hybrid design thus compensated for the soft and hard tissue deficit, which were difficult to fill with the pontic teeth alone, in addition to the extended span length. To the authors’ knowledge, these design combinations have not been examined and compared in the past with I- and TI-DC-RDPs; hence, the outcomes of this study could not be meaningfully compared with other similar papers with regards to the impact of design differences.

The 10-year tooth abutment survival rate (98.7%) in this study was higher than previously reported [1, 9, 14, 18]. The 10-year implant abutment survival rate (IaSR) was high in the present study. Previous studies reported IaSR in the range of 82.3% [5] to 97.4% [18], and a recent review revealed cumulative IaSRs for DC-RDPs (TI, 98.72%; I, 98.83%) [1]. Koller et al. [13] in their systematic reviewing survival of TI- and I-DC-RPDs reported a survival rate of 100% for teeth-implant-supported prostheses in the maxilla, as well as for tooth-supported RDPs. Furthermore, they reported survival rates of 97–100% for implant-supported prostheses in the mandible.

The slightly higher complication rate with Straumann implants compared with Camlog implants may be due to the more frequent placement of Straumann implants. No lost implant was replaced in an area with sinus augmentation. Technical complication rates were significantly higher for SAD TI-RDPs in the maxillary arch than for CAN prostheses in both arches; however, there were no statistical differences between the CAN and SAD I-RDPs. The variations may be related to the differential mobility between teeth and implant abutments, causing acrylic resin free-end SAD and veneer fractures [20].

The incidence of complications tended to increase over the first 10 years, but not significantly. More biological complications occurred in the group A (PE) with TI-RDPs than in the group B (CE) with I-RDPs. Occurrence of peri-implantitis has been reported, ranging between 4.3% and 10% [11, 15, 22‒24]. In our study, the higher occurrence of peri-implantitis in patients with I-RDPs than in those with TI-RDPs is in concordance with previous studies demonstrating patient-based peri-implantitis and implant loss more frequent in patients with I-DC-RDPs [9, 18, 25, 26]. Although limited data elucidates the reasons for higher peri-implantitis risk with I-RDPs, the preservation of natural teeth with TI-DC-RDPs may allow for the maintenance of functional oral proprioception, enabling wearers to limit stress loading on dentures and supporting implants and thereby reducing the risk of implant failure due to peri-implantitis [9]. The low rates of mucositis in this study may be attributed to the patients’ compliance to supportive periodontal care [3, 4, 18, 24, 27, 28]. Detailed descriptions of the occurrence of gingivitis and periodontitis in patients with DC-RDPs are scarce. Both conditions were observed more frequently in patients with TI-RDPs than in those with I-RDPs in this study. CE arches (group A) demonstrated similar rates of periodontitis 1 and 2. However, Rinke et al. [23] and Fritsch et al. [11] observed no periodontitis in patients with TI-RDPs [11, 20]. Recession occurred slightly more frequently in the CE arches (group A) than the PE ones (group B) at a median timepoint of about 6 years in this study. At all other follow-up timepoints, the 2 groups showed similar risk of recession development. Memon et al. [28] observed recession within 5 years in a majority of PE arches with conventional RDPs but not in arches with DC-RDPs and suggested it may be attributed to reduced retention and stability. In our study, fewer recessions were observed among patients with SAD I-RDPs compared with CAN I-RDPs and the reverse occurred for TI-RDPs.

The restoration survival rate (rSR) for groups of DC-RDPs was 100%. Schwarz et al. [29] observed rSRs of 93.3% and 100% (I-DC-RDPs and TI-DC-RDPs, respectively). Frisch et al. [11] reported rSRs for TI-DC-RDPs of 100% for a mean period of 6 years [17]. Rammelsberg et al. [9] observed 5-year rSRs of 85% and 92% (for I-DC-RDPs and TI-DC-RDPs, respectively). In this study, 5 (4.5%) DC-RDPs were repaired due to biological complications and 20 (1.8%) DC-RDPs were repaired due to technical complications. The DC-RDPs continued to function in the 3 cases of implant failure and in the 2 cases of tooth loss.

Although the present study had a large sample size, the distribution among different groups and subgroups was not uniform and may not have been adequate to elicit clinically significant differences between different groups. Second, information on the opposing arch/es has not been included in this study, and this factor might have affected the clinical outcomes. Furthermore, follow-up radiographic records could not be acquired for analyses due to country-specific radiological assessment regulations.

The 10- and 15-year abutment and prosthesis survival rates determined in this study indicate that I-DC-RDPs and TI-DC-RDPs could be recommended for restoration of PE and CE arches. Regarding the RDP design, CAN prostheses produced significantly fewer technical complications than SAD designs.

Required ethical approvals were obtained from the Dental Council North-Rhine, Germany (No.: RA 232.20 AK-cls), for conducting and publishing the results of this study.

The authors declare that there is no conflict of interest.

There were no funding resources for this work.

Gregor-Georg Zafiropoulos designed the study, performed all clinical procedures, and contributed to the writing of the manuscript. Moosa Abuzayeda performed statistical analysis and contributed to the interpretation of the results. Colin Alexander Murray selected and reviewed the scientific literature and edited the final manuscript. Mirza Rustum Baig contributed to the analysis and interpretation of the results and wrote the manuscript. All authors contributed equally to discussing the results and reviewing the manuscript before submission.

Data are not available for sharing.