Search strategy and eligibility criteria

This systematic review was performed in accordance with the PRISMA statement; the PICO(S) questions were used as evaluation criteria in order to identify the Patient or Population, Intervention, Control and Comparison, Outcome, and Study types.

A literature search was performed to identify available articles reporting on the clinical outcomes of Ti–Zr dental implants. A systematic approach was used to search the National Library of Medicine (Medline via PubMed) for articles published up to December 2014, including the following terms: ‘titanium–zirconium’ OR ‘Ti–Zr’ OR ‘Roxolid’. The electronic search was supplemented with a manual search of the following publications: International Journal of Periodontics and Restorative Dentistry , International Journal of Oral and Maxillofacial Implants , Journal of Periodontology , Implant Dentistry , Dentistry Today , Journal of Oral Implantology , Quintessence International , International Journal of Oral and Maxillofacial Surgery , Clinical Oral Implants Research , and Journal of Clinical Periodontology ( Fig. 1 ).

Flow diagram describing the search strategy.

The search resulted in a total of 162 hits from which eight abstracts were considered potentially relevant, while the manual search yielded two additional abstracts. Two reviewers (PA, EL) independently evaluated the abstracts against the inclusion and exclusion criteria, and the full-text articles were obtained. A third reviewer (JN) was consulted to confirm the eligibility of the selected articles.

Clinical (human) studies on Ti–Zr dental implants that fulfilled the following inclusion criteria were selected: (1) clinical studies of at least 10 treated patients; (2) prospective studies including randomized-controlled and non-randomized controlled studies and cohort studies; (3) retrospective studies including controlled studies, case–control studies, and single cohort studies; (4) a mean follow-up period of at least 6 months; (5) inclusion of data on the survival rate of the implants. The following exclusion criteria were applied: (1) articles written in languages other than English; (2) review articles; (3) studies with fewer than 10 patients, or case reports; (4) a mean follow-up period of less than 6 months. The level of agreement between reviewers regarding study inclusion was calculated using the kappa value.

Data extraction

Full text data extraction was performed independently for each eligible article by at least two reviewers (PA, EL). The following variables were extracted from each study: author(s), year of publication, study design, total number of patients, inclusion and exclusion criteria, follow-up duration, study outcomes (survival and success rates, marginal bone loss (MBL), and peri-implant measurements), patient demographics, implant type and manufacturer, total number of implants placed and number of implants in each patient, failed implants, jaw segment, bone regeneration needs, prosthetic complications, and loading protocols.

The methodological quality of the studies included was evaluated by one reviewer (PA) with regard to study design, randomization method, allocation concealment, blinding to the patient and examiner, and drop-out rates. In addition, the “Levels of evidence” document from the Centre for Evidence-Based Medicine was used to determine the qualitative validity.

Statistical analysis

Due to the heterogeneity of the articles, the difference in reported variables, and the inclusion of only two randomized controlled clinical trials, three objective outcomes could be extracted to perform the meta-analysis. The meta-analysis consisted of a weighted average for survival, success, and MBL, based on a random-effects model, with a confidence interval of 95%. The statistical analysis was performed using R 3.0.2 software (R Foundation for Statistical Computing, Vienna, Austria).

Types of studies

The selected studies included the following study types: two randomized controlled trials, six prospective studies, and one retrospective study. The studies reported cases of single-tooth gap, partial, or fully edentulous patients and implant placement in the anterior and posterior regions of both the maxilla and mandible ( Table 1 ).

Implants and demographic data

All implants placed in the studies included in this review were manufactured by the same company from a Ti–Zr alloy (Roxolid; Institut Straumann AG). Tissue-level and bone-level implants were used, with a reduced implant diameter of 3.3 mm. Bone-level implants had a conical shape and a narrow cross-fit connection. Tissue-level implants had a cylindrical shape and a 4.8-mm platform, with a machined collar design. In contrast to titanium NDIs, Roxolid NDIs are recommended for the restoration of anterior and posterior teeth.

A total of 607 patients received 922 narrow diameter Ti–Zr implants. Follow-up ranged from 3 to 36 months ( Table 1 ). The patients treated ranged in age from 21 to 76 years, and a higher proportion of women than men were treated in the studies.

Implant placement and loading protocols

Most of the studies followed early or delayed implant placement protocols according to the classification of Esposito et al. Two articles included immediate implant placement. Early or delayed loading protocols as defined by Weber et al. in 2009 were mostly used. Immediate loading was only described in three articles.

Clinical outcomes

For this review, the clinical outcomes of the nine studies were evaluated in terms of the survival and success rates and MBL. Only one of the nine studies did not present well-defined success criteria. Reported survival rates ranged from 94.7% to 100% and success rates from 94.7% to 100%. Three articles presented a sample size of 20 patients. One implant failed in each test group in two articles and two implants in the other, causing a drop in the survival rate.

Three studies reported mean pocket probing depths, with values of 2.69 ± 0.8 mm, 3.0 ± 0.74 mm, and 2.9 ± 1.2 mm. Another study presented a mean probing pocket depth that ranged from 2.21 to 2.89 mm after two years of loading. Mean MBL ranged from 0.16 ± 0.42 mm to 0.41 ± 0.56 mm at 1 year after implant placement, and from 0.33 ± 0.54 mm to 0.58 ± 0.60 mm at 2 years. Only one study reported MBL after 3 years of follow-up, with a value of 0.78 ± 0.75 mm ( Table 1 ).

After normalization of the results, a weighted average was obtained with a confidence interval of 95% for survival, success, and mean MBL. At 12 months after implant placement, the survival and success rates of the 901 implants were 98.4% and 97.8%, respectively, and the MBL reported for 156 implants was 0.36 ± 0.06 mm. At 24 months, the survival and success rates of 676 implants were 97.7% and 97.3%, respectively, and the MBL reported for 148 implants was 0.41 ± 0.09 mm ( Figs 2–7 ).

Survival rates after 1 year.

Survival rates after 2 years.

Success rates after 1 year.

Success rates after 2 years.

Mean marginal bone loss after 1 year.

Mean marginal bone loss after 2 years.

Surgical and prosthetic complications

Surgical complications described in these articles were limited to local inflammation in one patient. In the other studies, no surgical complications occurred. Quirynen et al. reported five cases of minor inflammation during the healing phase. No implant fractures occurred in the articles reviewed. The most frequently reported prosthetic complications in the study of Quirynen et al. were prosthesis fracture in nine cases (19%), loosening of a prosthetic component in three cases (6%), and prosthetic maintenance in three cases (6%). In another study, four patients (18.2%) had prosthetic complications such as abutment screw loosening. An absence of mobility and no need for prosthesis repair was reported in one study. There was a 100% prosthetic success rate in three studies. Prosthetic complications were not assessed in the other studies.

Bone regeneration needs

No bone regeneration was performed in three of the studies. Minor bone regeneration for small defects such as fenestrations or dehiscences was performed at the time of implant placement in two studies. In the study by Cordaro et al., bone regeneration was avoided in four out of 10 patients (40%) with the use of NDIs. Another study, using immediately loaded NDIs in partially edentulous patients, reported the same outcome. More than half of the patients (54%) avoided augmentation procedures in a non-interventional study. Finally, bone regeneration was not required in 11 out of 18 patients (61%) in a prospective study by Chiapasco et al.