The removal of deep carious lesions requires careful excavation to avoid pulp exposure and the need for endodontic therapy, and the stepwise or partial caries removal techniques have been indicated as conservative approaches to reducing these lesions.10, 12 Unlike in partial caries removal, in stepwise excavation, the clinician reopens the cavity and removes the remaining carious tissue with low risk of pulp exposure.10, 12 Despite the lack of evidence for the necessity of reentry to increase restoration success,11, 25 we used stepwise excavation in our study to evaluate the effects of provisional restorative materials on the remaining dentin. We hypothesized that simplifying the provisional restoration by using only a single light-cured material would not affect the outcomes of stepwise excavation. The results of our trial demonstrate that both procedures were able to improve the clinical aspects of the remaining dentin and to reduce the contamination of the tissue. Regarding the primary outcome, we observed no difference in the risk of developing necrosis or irreversible pulpitis between the protocols. Thus, we accepted the hypothesis of our study.

The occurrence of irreversible pulpitis and necrosis is affected strongly by any inflammation of pulp tissue and reduced thickness of the remaining dentin. We observed no difference in the thickness of the remaining dentin on the pulp floor at baseline between the cavities in the 2 treatment groups. Regarding pulp condition before the intervention, we included only carious lesions in teeth with a normal response to the cold test. The cold test presents high positive (100%) and negative (90%) predictive values when used in the diagnosis of pulp necrosis.26 However, cold sensitivity tests cannot be used to identify precisely the different degrees of pulp inflammation, and the absence of painful symptoms cannot be used to exclude any inflammatory involvement of the pulp.27 Thus, imprecision in the measurement of the previous condition of the pulp is a bias inherent to clinical trials involving restoration of deep carious lesions, and any relation between previous pulp condition and the primary outcome of our study is speculative. Patient age is another factor associated with differences in pulp response, given that older patients have reduced pulp response and are more at risk of developing irreversible pulpitis or necrosis.28 We included only patients aged 15 to 30 years in this study in an attempt to reduce the risk of bias related to age, and there were no differences between participants in the test and control treatment groups. Finally, the clinical and microbiological aspects of cavities that received different treatments were similar at baseline.

Considering that the baseline characteristics of participants and caries randomly assigned to receive control and test interventions were similar, it is reasonable to relate any differences in outcomes to properties of the materials. In our study, we reentered the tooth within a short period (3 months) after placing the provisional restoration. Despite the short period, we observed significant modifications in the scores for all clinical criteria and microbial counts. We selected the 3-month duration of the provisional restoration to increase study participant compliance and to facilitate follow-up care. On the basis of the primary outcome of pulp vitality according to the results of cold sensitivity testing, both treatments resulted in similar risks of developing irreversible pulpitis or pulp necrosis. In a meta-analysis, the investigators compared RMGI and calcium hydroxide placed in deep cavities and reported less inflammatory cell response with calcium hydroxide after 60 days; however, they observed no relevant clinical significance.29 This meta-analysis was based on a low level of evidence that had a high risk of bias, which indicates the need for well-conducted clinical trials. The results of our clinical trial indicated similar clinical outcomes whether we dressed deep cavities with calcium hydroxide or used RMGI alone. In this regard, RMGI causes mild initial pulp alterations when used in deep cavities; however, the inflammation subsides over time, providing evidence that the material is biocompatible.17
We observed modifications in dentin color and consistency and moisture on the cavity floor after removal of provisional restorations, irrespective of the restorative protocol used. The outer carious dentin tissue (infected dentin) consists of denatured collagen that does not have any potential for remineralization, whereas the collagen organization increases toward the inner, more mineralized dentin.30 We removed the outer infected dentin during the first excavation procedure and placed the provisional restorations over more organized dentin. Results from in vitro, in vivo, and clinical studies have provided indirect evidence that both RMGI and calcium hydroxide can remineralize carious dentin.12, 15, 16, 18, 29, 31 However, despite the evidence of ion exchange between these materials and dentin at a chemical level, the evidence to support interfibrillar or intrafibrillar remineralization of the carious dentin is weak.32, 33 Even in the absence of the true remineralization promoted by fluoride and calcium ions, the pH of calcium hydroxide and RMGI during the initial setting causes the release of transforming growth factor β from the dentin matrix, inducing odontoblast cells to produce tertiary dentin.34 Thus, increased dentin thickness on reentry could be expected, but we observed no difference, irrespective of the treatment used. One possible explanation for this finding is that the significant thickness of radiopaque dentin under carious lesions at baseline (on average, greater than 1.5 millimeters) impaired any effect of the calcium hydroxide and glass ionomer on the pulp tissue.

Provisional restorations reduced the count of Lactobacillus species and S mutans in the dentin. Proper sealing of the cavity alone is able to reduce the microbial count in carious dentin because of the limited amount and complexity of the available nutrients, regardless of the material used.35, 36 Furthermore, the high pH of calcium hydroxide (approximately 9.5) causes damage to organelles and cellular lysis,37 and the low pH of RMGI during the initial setting (approximately 4.0) affects the production of adenosine triphosphate and causes structural damage to the bacterial membrane because of the presence of hydrogen ions.38 It is important to emphasize that the dentin was contaminated despite the reduction in the microbiota after the placement of provisional restorations. However, the presence of bacteria under the restoration does not indicate the necessity of reentry to remove the carious tissue completely. There is no evidence that bacteria remaining under restorations led to progression of carious lesions.39
The results of our clinical trial demonstrate that the placement of a calcium hydroxide liner under provisional restorations during stepwise excavation does not improve the short-term outcomes when compared with the use of RMGI alone. For the data analyses, we followed the ITT approach and attributed the most unfavorable clinical outcomes (lack of pulp vitality and no alteration in secondary outcomes) to patients who did not return for follow-up. Therefore, we probably overestimated the adverse effects of partial caries removal on pulp vitality and underestimated the changes in secondary outcomes from the provisional restorations. Conversely, using ITT data analyses reduces the occurrence of unknown bias.24 A limitation of our clinical trial was the absence of follow-up of definitive restorations, given that a delayed pulp response can occur and should be evaluated in further studies.