Survival rates of resin composite restorations in loadbearing situations in posterior teeth

Investigating success rates of restorations

Randomized Clinical Controlled Trials (RCCTs) have been considered to be the ‘gold standard’ for assessment of the clinical performance of restorations by Brocklehurst and Hoare,⁵ these workers adding that ‘there are significant challenges when designing such complex studies’, and ‘there is an increasing need for high-quality evidence from primary dental care’. It has also been noted that RCCTs are few in restorative dentistry.⁶ In addition, the Cochrane review has been considered the ‘gold standard’ for systematic reviews, but these suffer from the small number of RCCTs in restorative dentistry. For example, in the most recent Cochrane study on posterior composites, published in 2014,⁷ the reviewers located seven RCCTs, but five were published in or before 1990 and the other two, while they were well designed RCCTs, were studies of amalgam vs composite in posterior teeth in 921 children under the age of 12 years. It may therefore be considered that these are not typical of the clinical situations where the majority of dental restorations for adult patients, worldwide, are placed, namely, general dental practice. The authors concluded that amalgam fillings had lower failure rates than composite fillings, but Burke⁶ considered that that conclusion cannot be justified because the research was on restorations in children and that the other papers were from a previous ‘generation’ of dental materials (bonding agents and resin composite). The authors add, rather strangely dismissing their findings, that ‘some materials may no longer be in use or have been replaced by products with better physical properties’.

It was therefore suggested that future Cochrane reviews and, indeed, all future systematic reviews on restoration survival in restorative dentistry, should include well-designed cohort studies which evaluate large numbers of restorations, and/or long duration, simply because there are more of these than well-designed RCCTs.⁶

This narrative review will therefore seek to identify broadly-based studies in which amalgam and resin composite restorations for adult patients are compared over a minimum period of five years and will also include cohort studies for adult patients, where possible from general dental practice, in which more than 250 restorations are followed for a minimum period of five years (Table 1). It should be added that success with ‘posterior composites’ is dependent also upon the operator's knowledge of, and familiarity with, the various technique sensitivities which have been described by Mackenzie and colleagues.⁸

Methods

An OVID Medline and Epub Ahead of Print Search was carried out with the title ‘Survival of composite restorations in posterior teeth’ – English language articles 1990–2018. In addition, the references from selected papers were searched in order to ensure as complete a coverage of the subject as possible.

Studies including amalgam and composite restorations

It is not possible to compare the results of the above studies directly due to their heterogeneity, but the conclusion which may be gleaned is that resin composite restorations have survival rates similar to amalgam when placed in loadbearing situations in posterior teeth, with composite being superior to amalgam in two studies, worse in two and with four of the papers which directly compared resin composite with amalgam being equivocal in their findings. There is obviously a need for a RCCT examining survival of amalgam and resin composite but the authors are dubious that this will ever happen.

Composite-only studies

The following studies have been identified:

Results

A total of 144 studies fulfilled the initial search criteria. However, when the results were re-assessed with regard to the inclusion criteria, studies relating to primary teeth, ART, CAD-CAM, composite inlays/onlays, fibre-reinforced composite, resin-modified glass ionomer (RMGI), composite crowns, sealant restorations, preventive resin restorations, and core build-ups were excluded. The number of studies finally included was 24.

Where the data permit, the authors will calculate the Annual Failure Rate (AFR), as a readily understandable and comparable means of determining survival of restorations, this being the overall failure rate over the time of the particular study divided by the number of years for which the restorations were assessed.

The conclusion gleaned from the above cohort studies is that resin composite restorations have acceptable survival rates when placed in loadbearing situations in posterior teeth, with AFRs generally within the range 2% to 3%, which the authors consider to be compatible with successful clinical practice. However, the relevance of the presence of a glass-ionomer base is not clear as it does not appear to increase the AFR in all the identified studies. This may be a matter for individual clinicians to consider, based upon the risk factors of their patients.

Systematic reviews

The conclusion gleaned from the above systematic reviews is that resin composite restorations have acceptable survival rates when placed in loadbearing situations in posterior teeth, with AFRs generally within the range 2% to 3%. Risk factors for premature failure include patients at high risk of caries and the presence of a liner or base beneath the resin composite restoration.

Discussion

This narrative review has sought to identify publications which can provide meaningful information about the survival of posterior composites, as these restorations are being increasingly placed, worldwide. A high proportion of the included cohort studies on posterior composite have involved restorations placed in general dental practice, with the authors considering that this may enhance the relevance of the findings. In addition, three of the papers which directly compared amalgam with composite were also set in the general dental practice arena. While the inclusion criteria stated ‘adult patients’, two studies have been included^{22, 23} in which the posterior composite restorations were placed in adolescents. The authors consider that these are relevant, firstly because of the large numbers of restorations assessed at over five years in a primary care environment, but also because it may be considered that, if restorations can survive in this potentially challenging group of patients (for factors such as high caries diet and challenging patient co-operation), then they should perform at least as well in older patients.

It may therefore be concluded that posterior composite restorations may provide restorations of satisfactory survival, with two papers^{17, 34} of those identified in which amalgam and composite are directly compared, having posterior composite restorations which performed less well than amalgam. However, the authors consider that the number of amalgam restorations in one study¹⁷was so small as to preclude meaningful analysis while, in the other,³⁴the authors advised that the results should be interpreted with caution.

The papers which were selected were considered to be those which had relevance to clinical practice but which presented large numbers of restorations, assessed over periods of more than five years. In this regard, survival of restorations at one year only produce ‘evidence’ relating to the catastrophic failure of restorations, and the systematic review by Opdam et al has demonstrated that long-term data are essential.³⁰

Annual failure rates of posterior composite restorations may be considered to provide restorations of satisfactory longevity. Risk factors, such as high caries susceptibility, large restorations (particularly in molars) have been particularly identified. In the present study, posterior composite restorations were found to have an AFR of between 1.3% and 3.0%, which the authors consider to be compatible with successful clinical practice. However, it is interesting to note that Alvanforoush and colleagues³⁶ did not identify any improvement in performance of posterior composite restorations during the two time periods, 1995–2005 and 2006–2016, despite what many clinicians, anecdotally, might have considered improvements in the materials that they use. The overall rates of failure showed little difference, but the causes showed a notable change. This is believed to be a reflection of increased use of composites for larger restorations and possibly changes of material characteristics. This has also been observed in analysis of a 10 million dataset of restorations placed in England and Wales, with no improvement noted in any form of restoration between 1990 and 2006.³⁷

It may be considered that a majority of the restorations placed in the included studies used traditional cavities which included an occlusal lock/key, rather than the mini-saucer shaped cavities (Figure 2), which may be considered appropriate when using adhesive techniques. There is, however, a paucity of work on such cavities although, in the paper by Pallesen and colleagues,²³ they state that the cavity design utilized was that which allowed minimal removal of tooth substance, as was appropriate to the ingress of caries. Those workers reported an overall AFR of 2.0%, so it may be considered that such cavity designs are appropriate. On the other hand, Kopperud and co-workers¹⁷ considered that ‘saucer-shaped cavity preparations appeared to predispose to reduced survival of the resin composite restorations’. However, the early work of Nordbo et al,³⁸ who followed saucer-shaped cavities restored with resin composite, concluded that ‘the saucer-shaped resin composite restoration represents a viable treatment modality for small cavities, and that the time may have come to include it in dental curricula as a routine operative treatment for small Class II lesions’, a positive endorsement. Further research is therefore indicated on these minimal cavity designs and, in particular, their effect on the fracture resistance of teeth, especially premolars when compared with traditional cavity designs.

However, despite their successful survival data, as identified in this review, posterior composites may not be cost-effective, because they take longer to place.¹ In this regard, Tobi and colleagues³⁹ assessed this, as part of a larger randomized clinical trial, looking at effectiveness and treatment costs for 73 composite and amalgam Class II posterior restorations. The results indicated that replacing an amalgam Class II restoration with another amalgam is associated with lower costs than replacing with a composite resin. However, the difference in cost-effectiveness between posterior composites and amalgam restorations may be minimized when using new bulk fill restorative materials which may be placed in increments of up to 5 mm, by recently-introduced adhesives which are quick to place and by increasing experience and skill of practitioners using these materials. In this regard, results from a recent research abstract⁴⁰ from the London meeting of the International Association for Dental Research have indicated that bulk fill composites may indeed be placed faster than conventional composites in posterior teeth.

While it has been considered that the RCCT is the ‘gold standard’ for comparing two different interventions, these are few, hence the dependence upon cohort studies and systematic reviews in this review. A group of renowned experts in restorative dentistry, co-ordinated by Reinhart Hickel, have published recommendations for the conduct of controlled clinical studies of restorative dental materials.⁴¹ They stated that ‘there are many publications on clinical studies but, in most cases, it is impossible to analyse and compare results. This is largely due to inadequate design, insufficient reporting on restoration placements, inadmissible pooling of different groups'. While all of this is true, this study has attempted to calculate AFRs as a readily understandable means of understanding whether restorations are successful or not.

Clinicians reading the present work will all have an aspiration as to the longevity of restorations that they place on a daily basis, so will be able to understand whether AFRs published in the literature are satisfactory or not. In this regard, while it is essential that clinicians have knowledge of the research publications relating to the treatment which they are carrying out, data from the clinician's own clinical practice may be considered valuable when communicating with patients, who may better understand survival rates quoted by their dentist from his/her own practice as opposed to the research literature. In the first author's practice, he followed up many of the restorations which he had placed at periods of up to 10 years (examples being presented in Figures 3, 4 and 5), building up an objective assessment of how well (or not) his restorations were performing. However, it is suggested that this is not good enough today, and that clinicians should be encouraged to adapt the computer system in their practice to record survival data, this surely being the ultimate in meaningful data when discussing possible treatments with their patients.

Finally, the authors are aware that some colleagues still have significant problems with posterior composites. It is suggested that, factors influencing longevity of posterior composites are many, but could include those listed below:

Conclusions

The conclusion, based upon this review, is that posterior composites may provide restorations of satisfactory longevity and with survival rates generally similar to those published on amalgam restorations, previously considered as the ‘gold standard’. This review has demonstrated that there is clinical evidence to back up these statements. However, the ability of the operator in placing the restoration may have a profound effect, as several studies have demonstrated. In this regard, practitioners who have not had experience in placing posterior composite restorations in dental school need to be prepared to attend practical continuing education courses if they are not convinced about their skills and success rates with posterior composites, particularly in restoring extensive cavities.