CAD-CAM removable complete dentures: A systematic review and meta-analysis of trueness of fit, biocompatibility, mechanical properties, surface characteristics, color stability, time-cost analysis, clinical and patient-reported outcomes

: Objectives This review compared Computer-aided designand Computer-aided manufactured (CAD-CAM) and conventionally constructed removable complete dentures (CDs). Data Seventy-three studies reporting on CAD-CAM (milled/3D-printed) CDs were included in this review. The most recent literature search was performed on 15/03/2021. Sources Two investigators searched electronic databases [PubMed (MEDLINE), Embase, CENTRAL], online search engines (Google) and research portals. Hand searches were performed to identify literature not available online. Study selection Studies on CAD-CAM CDs were included if they reported on trueness of fit, biocompatibility, mechanical, surface, chemical, color , microbiological properties, time-cost analysis, and clinical outcomes. Inter-investigator reliability was assessed using kappa scores. Meta-analyses were performed on the extracted data . Results The kappa score ranged between 0.897–1.000. Meta-analyses revealed that 3D-printed CDs were more true than conventional CDs (p = 0.039). Milled CDs had a higher flexural-strength than conventional and 3D-printed CDs (p < 0.0001). Milled CDs had a higher flexural-modulus than 3D-printed CDs (p < 0.0001). Milled CDs had a higher yield-strength than injection-molded (p = 0.004), and 3D-printed CDs (p = 0.001). Milled CDs had superior toughness (p < 0.0001) and surface roughness characteristics (p < 0.0001) than other CDs . Rapidly-prototyped CDs displayed poor color-stability compared to other CDs (p = 0.029). strain at yield point than milled CDs superior esthetics than 3D-printed (p 0.0001). of CAD-CAM CDs chairside time (p = and costs (p 0.0001) than CDs. systematic review concludes that CAD-CAM CDs offer a number of improved mechanical/surface properties and are not inferior when compared to conventional CDs. Clinical significance CAD-CAM CDs should be considered for edentulous whenever this numerous advantages retention, and surface preserves a Studies on CAD-CAM CDs were included if they reported on trueness of fit, biocompatibility, mechanical, surface, chemical, color , microbiological properties, time-cost analysis, and clinical outcomes. Inter- investigator reliability was assessed using kappa scores. Meta-analyses were performed on the extracted data . Results: The kappa score ranged between 0.897 – 1.000. Meta-analyses revealed that 3D-printed CDs were more true than conventional CDs (p = 0.039). Milled CDs had a higher flexural-strength than conventional and 3D-printed CDs (p < 0.0001). Milled CDs had a higher flexural-modulus than 3D-printed CDs (p < 0.0001). Mil- led CDs had a higher yield-strength than injection-molded (p = 0.004), and 3D-printed CDs (p = 0.001). Milled CDs had superior toughness (p < 0.0001) and surface roughness characteristics (p < 0.0001) than other CDs . Rapidly-prototyped CDs displayed poor color-stability compared to other CDs (p = 0.029). CAD-CAM CDs d displayed better retention than conventional CDs (p = 0.015). Conventional CDs had a higher strain at yield point than milled CDs (p < 0.0001), and had superior esthetics than 3D-printed (p < 0.0001). Fabrication of CAD-CAM CDs required less chairside time (p = 0.037) and lower overall costs (p < 0.0001) than conventional CDs. Conclusions: This systematic review concludes that CAD-CAM CDs offer a number of improved mechanical/ surface properties and are not inferior when compared to conventional CDs. Clinical significance: CAD-CAM CDs should be considered for completely edentulous patients whenever possible, since this technique offers numerous advantages including better retention, mechanical and surface properties but most importantly preserves a digital record. This can be a great advantage for older adults with limited access to dental care.

PICO focused question, criteria for inclusion, sources of information, search terms, search strategy, search filters, and search dates.

Focus question
In completely edentulous patients, are CAD-CAM removable complete dentures (CDs) inferior to conventional CDs with respect to trueness of fit, biocompatibility, mechanical properties, surface characteristics, color stability, time-cost efficiency, clinical and patient-reported outcomes? Criteria Inclusion criteria Studies reporting on CDs manufactured by CAD-CAM (milled/3D-printed) and conventional processesAll study designs complete denture OR removable complete denture OR removable partial denture OR removable dental prosthesis OR complete denture prosthetic OR complete denture prosthodontics OR diagnostic denture OR immediate denture OR provisional denture OR transitional denture OR treatment denture OR trial denture OR full denture OR interim denture OR interim prosthesis OR overlay denture OR digital workflow OR implant supported removable dental prostheses OR implant supported complete removable dental prosthesis OR implant supported partial removable dental prosthesis OR implant supported overdenture OR implant assisted over dentures NOT implant fixed Comparison #3 #3. 1 CAD CAM denture OR Computer Aided Design denture OR Computer Aided Manufacturing denture OR Computer Assisted Machining denture OR CNC denture OR Computer Numerical Control denture OR digital denture OR digitally fabricated denture OR CAE denture OR Computer Aided Engineering denture OR Milling CAD CAM OR 3D printed denture OR Milled denture OR subtractive fabrication denture OR three dimensional printed denture OR Stereolithography denture OR SLA denture OR additive fabrication denture OR rapid prototyping denture OR selective laser sintering denture OR additive layer denture OR DMLS denture OR Direct metal laser sintering denture OR SLS denture OR selective laser sintering denture OR Photo solidification OR resin printing Outcome #4 #4. 1    The fabricated dentures were hydrothermally aged and microwave sterilized. The trueness was measured before and after the aging process, using a 3D comparison software (Geomagic Control X, 3D Systems).
Before the aging process, the milled group demonstrated the lowest surface deviation, followed by the injectionmolded and 3D-printed groups. Hydrothermal cycling did not affect the milled group's trueness in contrast to the injection-molded and 3Dprinted groups. Microwave sterilization caused no effect on the 3D-printed group's dimensional trueness; but led to clinically critical deformations of the injectionmolded and milled groups.     Bonded, the denture teeth were bonded into the milled base; DLP, digital light processing; FDM, fused deposition modeling; Monolithic, the denture teeth were milled as part of the denture base; n, sample size; NS, not specified; PLA, polylactic acid; W/Tray, copy denture technique with tray; SD, standard deviation; SLA, stereolithography; †, study used for meta-analysis.

Introduction
Epidemiological surveys indicate that people are both living longer and retaining more of their natural teeth into old age. [1][2][3]. Rehabilitation of completely edentulous jaws with conventional removable complete dentures (CDs) is a well-established treatment protocol. Traditionally, CDs are fabricated either as a completely new CD or by using copy techniques [4][5][6]. Whilst some clinical aspects of these techniques differ, they both include intra-oral impressions taken of the denture bearing areas with occlusal information provided using wax rims. However, the use of computer-aided design and computer-aided manufacturing (CAD-CAM) techniques in the construction of CDs has recently gained popularity [7]. CAD-CAM CDs can be constructed in as few as two clinical visits. At the first visit, all clinical records are captured, which can take the form of traditional impressions or digital records produced using intra-oral scanning technology. The records are transferred to the digital dental laboratory, where the entire denture is designed virtually. A design preview for the clinician to approve is possible for some techniques, before the digital dental laboratory completes the denture. At the second clinical visit, the dentures are ready for insertion. Whilst this technology is still in its infancy, it may offer significant benefits to older patients, including fewer clinical appointments alongside some reports of improved fit and better material properties compared to traditionally manufactured dentures [8].
Despite the increasing availability of CAD-CAM CDs, the majority of edentulous patients still receive dentures constructed using more traditional techniques. In this review, conventional techniques employed to fabricate CDs include flask-pack-press (FPP) or injectionmolding using polymethylmethacrylate (PMMA) resin materials that may be either heat-polymerized or auto-polymerized, polyamides, or composite resin materials. In comparison, the CAD-CAM methods referred to are either additive [rapidly-prototyped (RP)/3D-printed] or subtractive (milled) processes. The 3D-printing techniques include stereolithography, digital light processing or fused deposition modeling. This aim of this systematic review was to evaluate and compare CAD-CAM CDs with conventionally manufactured CDs in terms of trueness of fit, biocompatibility, mechanical properties, surface characteristics, color stability, time-cost analysis, clinical and patient-reported outcomes. The PICO (Population Intervention/exposure Comparison Outcome) focused research question set for this systematic review was: "In completely edentulous patients, are CAD-CAM removable complete dentures (CDs) inferior to conventional CDs with respect to trueness of fit, biocompatibility, mechanical properties, surface characteristics, color stability, time-cost efficiency, clinical and patient-reported outcomes?"

Protocol and registration
This systematic review was conducted and reported according to the PRISMA (preferred reporting items for systematic reviews and meta-  The flexural modulus of the CAD-CAM milled group was significantly higher than that of the other tested groups. i, manufacturer-recommended 3D-printer; ii, third-party 3D-printer; iii, printed in a vertical orientation; n, sample size; NS, not specified; SD, standard deviation; †, study used for meta-analysis. analysis) guidelines [9]. The protocol used in this systematic review is similar to the design used in previously published systematic reviews [10,11]. The review protocol was registered with PROSPERO: International prospective register of systematic reviews (CRD42020175673).

Eligibility criteria and information sources
The predefined list of inclusion and exclusion criteria used for this systematic review are detailed in Table 1. All studies reporting on CDs manufactured using CAD-CAM and conventional processes in completely edentulous patients were searched using online electronic databases (PubMed, Embase and CENTRAL). Relevant publications identified but which were not accessible online were hand-searched. Other sources such as online search engines (including Google Scholar and Yahoo), online research community websites (https://www. researchgate.net/), and reference cross-checks were all accessed to ensure the maximum pool of relevant studies was generated. No further searches were performed after the last update, which was on March 15th, 2021. i, manufacturer-recommended 3D-printer; ii, third-party 3D-printer; iii, printed in a vertical orientation; n, sample size; NS, not specified; SD, standard deviation; †, study used for meta-analysis.

Table 6
Studies reporting strain at yield-point of denture bases. i, manufacturer-recommended 3D-printer; ii, third-party 3D-printer; iii, printed in a vertical orientation; n, sample size; NS, not specified; SD, standard deviation; †, study used for meta-analysis. i, manufacturer-recommended 3D-printer; ii, third-party 3D-printer; iii, printed in a vertical orientation; n, sample size; NS, not specified; SD, standard deviation; †, study used for meta-analysis. CAD-CAM was not generally found to be better in fracture toughness than the conventionally manufactured groups. One of the six milled groups had a higher fracture toughness than the compression-molded group, while three of the six milled groups had a lower fracture toughness than the compression-molded group. Three milled groups had a higher fracture toughness than the autopolymerization group, while one of six milling groups had a lower fracture toughness than the autopolymerization group. i, without light-curing topcoat; ii, with light-curing topcoat; KIc, plane strain fracture toughness; n, sample size; NS, not specified; SD, standard deviation; †, study used for meta-analysis.

Search strategy and study selection
An initial search strategy was designed and set up by the investigators. Two investigators performed the searches based on the identified medical subject headings (MeSH) search terms as dictated by the search design and strategy. The terms were then applied using the appropriate Boolean operators, "OR" or "AND," or "NOT" to perform the search in the databases. The full set of search terms used and the filters set when performing the search in the above databases are described in Table 1. No restrictions were applied to the type of studies included. The investigators (PK and MS) initially swept through the search results using a thorough title and abstract screening. After the initial sweep, the shortlisted studies were included for a full-text analysis only after a mutual agreement between the two investigators. Disagreements, if present, were resolved through a consensus meeting. If multiple publications existed on the same cohort by the same author, only the most recent publication was included in the review.

Data collection process and missing data
Data extraction was performed independently by two investigators (PK and MS), who were reciprocally blinded to the each other's data extraction. The corresponding authors of the included publications were contacted by email for any clarification of extracted data from their studies. The parameters extracted from the included studies are detailed in Tables 2-20. For any missing information from the included studies relevant to this systematic review, direct email contact was made with the corresponding author. Email reminders were sent to the authors in case of a non response. Follow-up emails were sent if the received information was inadequate or required further clarity. A non response from the author ultimately lead to the exclusion of the study, when necessary information was lacking.

Summary measures and synthesis of results
Inter-investigator reliability was assessed using kappa (κ) statistics. The meta-analysis was performed comparing CDs manufactured using CAD-CAM and traditional processes with regard to trueness of fit, biocompatibility, retention, flexural strength, flexural modulus, yield strength, strain at yield point, toughness, fracture toughness, hardness, surface wettability, surface roughness, color stability, residual monomer content, clinical and patient reported outcomes. In this review individual subgroups in the studies were considered independent. For each of the studied parameters, means, standard deviations along with sample sizes were extracted. Confidence intervals (CI) were set to 95%, and standardized mean differences were calculated for each outcome parameter using comprehensive meta-analysis software, version 3.0 (Biostat, Englewood, NJ, USA). Random-effects or fixed-effects models were used to calculate the weighted means across the studies [12] and I-squared statistics (I 2 -statistics) was used to assess the heterogeneity across the included studies.

Risk of publication bias and additional analyses
Risk of publication bias was assessed across the studies using funnel plots [13]. Descriptive analysis was performed on all studies to report their outcomes, sample sizes, methods, conclusions as well as the fabrication techniques including brand and manufacturer names of sample materials used in each study.
were included after reference searching and hand searches to leave a final shortlist of 73 articles [8,. The flow of the entire systematic search and article identification process is illustrated in Fig. 1. The various CD processing techniques identified in this review has been summarized in Fig. 2. The overall κ scores calculated for the various parameters extracted by the two investigators ranged between 0.897 and 1.000, hence indicating an excellent degree of inter-investigator agreement.
All 73 publications in the final shortlist were analyzed and extracted data included outcome values, sample size, method, conclusions as well as the fabrication technique including brand and manufacturer of materials used in each study. The studies were categorized according to their measured outcomes as follows: trueness of fit, bonding ability to other materials, flexural strength, flexural modulus, elastic modulus, yield strength, strain at yield point, toughness, fracture toughness, hardness, surface wettability, surface roughness, color stability, biocompatibility, microbial adhesion (Candida albicans), residual monomer content, treatment time or cost, retention, esthetics, clinical outcomes and patient-related outcomes.
There were no significant differences in the number of unscheduled post-insertion visits for participants whose dentures were fabricated following injection-molding or milled protocols.  (Fig. 8, Table 6).

Color stability
A series of meta-analyses were undertaken to compare the color stability data for milled CDs; conventional (flask-pack-press) CDs;

15
Predoctoral dental students delivered two sets of dentures, compression molding and milling dentures, for each patient. Experienced and certified prosthodontists assessed denture quality. The students and patients were asked to complete the questionnaires.
The average clinical chair time was 205 mins longer for the compression-molded group than for the milled group. According to clinical outcomes, significantly higher average scores were observed for milling dentures than for compression-molded dentures. Both students and patients preferred milled dentures more than compression-molded dentures.
Milled Avadent, Global Dental Science 15 †Inokoshi et al. (2012) [24] Nonrandomized crossover trial Conventional wax trial denture NS Clinical outcomes (reported by prosthodontists) such as esthetics, stability, overall satisfaction. Patient-centered outcomes (reported by patients) such as esthetics, predictability of final denture shape, stability, comfort of the dentures, overall satisfaction.

10
Prosthodontists performed a denture try-in for one patient using both trial dentures from conventional and 3D-printing methods. The prosthodontists and patients rated satisfaction for both methods using a visual analog scale; VAS.
Regarding prosthodontist's ratings, esthetics and stability were rated significantly higher with the conventional method than with the 3D-printing method, whereas chair time was rated significantly longer with the 3D-printing method than with the conventional method. 3D-printing FullCure720, Objet Geometries 10 n, sample size; NS, not specified; †, study used for meta-analysis. Undergraduate final-year dental students utilized both the digital denture protocol and the conventional complete denture protocol to construct two sets of complete dentures for patients. Overall time spent and costs (clinical, materials, and laboratory) were calculated.

Descriptive analysis and quality assessment of the included clinical studies
Parameters where a meta-analysis was not possible were reported descriptively. Elastic-modulus, biocompatibility, anti-microbial adhesion, and the bonding ability of the CAD-CAM resins are reported descriptively in Tables 17-20. The characteristics of all the included studies, including all the extracted data, the outcome variables, sample sizes, methods, conclusions as well as the fabrication techniques enlisting the brand and manufacturer of materials are presented in the tables. A quality assessment of the included clinical studies was performed using the Newcastle-Ottawa scale for assessing the quality of non-randomized studies and is reported in Table 21.
Good adaptation of the denture base to the denture bearing tissues is essential for the adequate retention and stability of any CD [38]. Trueness of fit refers to the closeness of agreement between the expectation of a measurement result and a true value [85]. This review demonstrated that the trueness of fit for milled CDs was not significantly different from conventional, 3D-printed and injection-molded CDs, all techniques led to an clinically acceptable trueness of the intaglio surface. The clinical retention of a CD depends, apart from the morphology and resilience of the denture bearing tissues, on adaptation of the intaglio surface to the tissues, border seal, and salivary flow-related effects associated with viscosity and film thickness of the oral fluid [86,87]. Deformation of conventional denture body during processing is affected by the shape (palatal vault and residual ridge), thickness, denture base materials, and denture processing steps [88,89]. Mucosal adaptation, which is associated with retention, stability, and support, is influenced by distortion [49], hence all attempts to minimize distortion must be made. In conventional fabrication techniques, the deformation of heat-polymerized resin may diminish the degree of base adaptation This clinical misfit is being compensated by deliberately compressing the posterior palatal seal area and hence creating a suction effect, as well as a primary remount of the denture to correct the occlusal discrepancies which result from the denture deformation through polymerization.
Given the data on trueness of fit, this review also examined the issue of clinical denture retention. It is widely reported that successful CD therapy requires satisfactory stability, support and retention [90]. For conventional CDs posterior palatal seal design, palatal surface design, denture base surface enhancement and adhesives contribute to denture retention [91,92]. In the long term, denture wearing in neurologically healthy patients these parameters might be complemented or compensated by muscular skills. However, polymerization shrinkage of conventional CD bases can negatively impact on adaptation and retention [93]. This review demonstrated that the retention of CAD-CAM CDs was superior to conventional (flask-pack-press) (p = 0.015) CDs.
Data on a large number of mechanical properties were examined in this review. From the data analyzed CAD-CAM CDs exhibited superior performance in flexural strength; flexural modulus; yield strength; toughness; and surface roughness.
Hardness is a measure of the resistance to localized plastic deformation induced by either mechanical indentation or abrasion. CDs made of a material with low surface hardness can be damaged by mechanical brushing, causing plaque retention and pigmentations, which can decrease the life of the prostheses. Conventional CD bases are prone to  fracture, particularly with impacts sustained when the denture is out of the mouth or while in service in the mouth due to flexural fatigue as the base undergoes cyclical loading during mastication [94,95]. High flexural strength is imperative for sustained successful CD wear as alveolar resorption is a continual and irregular process which can lead to uneven prosthesis support [96]. To ensure that the stresses applied during mastication do not cause permanent deformation, wear and ultimately fracture, the CD base material must exhibit a high elastic modulus. A number of properties are responsible for microbial colonization of denture bases including surface roughness. Microbial contamination of denture surfaces can elicit localized intra-oral mucosal infections but have also been implicated in the aetiology of aspiration pneumonia in dependent older adults [97]. The surface roughness of conventional CD bases is primarily determined by processing which gives rise to gaseous porosities and surface irregularities. Although these irregularities can be countered by applying packing pressure, the amount of applicable pressure is limited in conventional CD manufacturing, as too high pressure may cause fractures of the mold or the flask [98,99]. By contrast, in CAD/CAM CD manufacturing, the bases are milled from industrially polymerized resin pucks, and the resin in these pucks is highly condensed because of the high pressure the manufacturers apply during polymerization. As illustrated in this review, the fully automated    . 4. Forest plot comparing the trueness of fit (mean and SD in mm) between 3D-printed (3DP), conventional flask-pack-press (C_FPP), injection-molded (C_injection), and milled, complete dentures. CI, confidence interval; 3DP_DLP, 3D-printed using digital light processing (3DP); 3DP_FDM, 3D-printed using fused deposition modeling (3DP); Hor, 3D-printed in horizontal orientation; Man, mandibular denture fabrication; Max, maxillary denture fabrication; 3DP, 3D-printed; SD, standard deviation; SLA, stereolithography (3DP); Std., standardized; Ver, 3D-printed in vertical orientation  . 6. Forest plot comparing the flexural modulus (mean and SD in MPa) of milled, conventional flask-pack-press (C_FPP), injection-molded (C_injection) and 3Dprinted (3DP) complete dentures. AVD, 'Avadent' (milled); CI, confidence interval; Mb, 'AvaDent Denture base disc' (milled); Ms, 'Avadent Extreme CAD-CAM shaded disc YW10' (milled); Rc, 'NextDent C&B' (3DP); Rm, 'NextDent Base' 3D-printed using a manufacturer-recommended 3D-printer (3DP); Rt, 'NextDent Base' 3D-printed using a third-party 3D-printer (3DP); Rv, 'NextDent Base' 3D-printed using a vertical orientation (3DP); SD, standard deviation; Std., standardized; TIZ, 'Tizian' (milled).
The articles identified in this systematic review did not include an extensive number of studies which utilized patient reported outcome measures (PROMs). Unfortunately, this is a common finding across removable prosthodontics and should be addressed in future research. Data was summarized on esthetics which was gathered from a series of Visual Analogue Scales (VAS) completed by clinicians. These results indicated that clinicians preferred conventional CDs in terms of esthetics (p = 0.002). When the CAD-CAM milled base was used in conjunction with conventional artificial teeth, no significance was noted between milled and injection-molded dentures [22]. However, when comparing conventional (flask-pack-press) and 3D-printed CD groups, a clear preference was found for the conventional (flask-pack-press) group [24]. It would appear that limited esthetics continue to be an issue with CAD-CAM CDs with patients expressing concern about the pink and white esthetics of the prostheses [24,100]. This issue should also be  Fig. 11. Forest plot comparing the hardness (mean and SD in MPa) between milled, 3D-printed (3DP), conventional flask-pack-press (C_FPP), and injection-molded (C_Injection) complete dentures. AVD, 'Avadent' (milled); CI, confidence interval; Mb, 'AvaDent Denture base disc' (milled); Ms, 'Avadent Extreme CAD-CAM shaded disc YW10' (milled); Rc, 'NextDent C&B' (3DP); Rm, 'NextDent Base' 3D-printed using a manufacturer-recommended 3D-printer (3DP); Rt, 'NextDent Base' 3Dprinted using a third-party 3D-printer (3DP); Rv, 'NextDent Base' 3D-printed using vertical orientation (3DP); SD, standard deviation; Std., standardized; TIZ, 'Tizian' (milled). considered in relation to the highly aesthetic conventional CDs which can be produced by high quality dental technicians particularly when working closely with both the clinician and the patient [101][102][103]. However, it is highly likely that the esthetics of CAD-CAM CDs will evolve rapidly in future with constantly improving technology.

Fig. 20.
Forest plot comparing the chair-side time (mean and SD in minutes) involved in fabricating conventional flask-pack-press (C_FPP), and milled complete dentures. CI, confidence interval; SD, standard deviation; Std., standardized; U, upper complete denture fabrication; U/L, upper and lower complete denture fabrication. serviceability. Unfortunately, there is an extremely small number of clinical studies which have utilized validated PROMs. Given that successful CD therapy is often built on a positive relationship between patient and clinician, incorporating the patient's opinions into the final prostheses, is very important [104]. This review did not identify any clinical studies which utilized Quality of Life measures, despite a number of instruments specifically developed for edentate older adults [105]. This should be addressed in future clinical studies with appropriate long-term follow-up. The majority of the studies included in this review were in vitro studies; currently, a universal methodological assessment tool for in vitro studies that assesses all critical aspects of in vitro metanalysis does not exist [106], hence quality assessment of these in vitro studies could not be performed. It is also important to mention the heterogeneity of the included studies, which may be considered a further limitation of this review. Although these limitations might have impacted the findings of this review, the methodology of this review adhered to all the recommended protocols for performing systematic reviews and therefore may be considered robust.

Conclusions
The introduction of CAD-CAM CDs has brought many advantages including fewer patient appointments, reduced clinical time and digital archiving of completed prostheses. Some CAD-CAM techniques also result in reduced manufacturing costs. This systematic review concludes that CAD-CAM CDs offer a number of improved mechanical/surface properties and are not inferior when compared to conventional CDs. However, further long-term follow-up studies are required to fully evaluate these CAD-CAM CDs with particular regard to estheticsand PROMs .

Source of funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflicts of interest Statement
The authors declare that they have no conflict of interests.

Table 21
The Newcastle-Ottawa Scale for assessing the quality of non-randomized studies.