Pulpotomy versus pulpectomy in the treatment of vital pulp exposure in primary incisors. A systematic review and meta-analysis.

Background: Early childhood caries is a serious public health problem. When caries extend to involve the pulp, various forms of pulp treatment are tried to stimulate tooth repair. Although pulpotomy is the treatment of choice for vital primary tooth pulp exposure but there is a trend among many dentists to perform pulpectomies in vital primary incisors. This study aimed to assess the effect of pulpotomy and pulpectomy in treatment of carious vital pulp exposure in primary incisors. Methods: We searched Pubmed and Cochrane library databases up to March, 2018, OpenGrey for grey literature and ClinicalTrials.gov for ongoing trials. Randomized controlled trials were included and assessed with Cochrane risk of bias tool . Primary outcomes were clinical failure and radiological failure. The effect sizes were calculated as risk ratios with 95%CI using the Mantel-Haenszel method. Results: Four trials were identified for qualitative assessment, only three trials were included in meta-analysis after exclusion of one trial due to its high risk of bias. The pooled results of the longest follow up period for clinical failure showed no statistically significant difference between pulpotomy and pulpectomy. The relative risk (RR) was e 2.69, 95% CI 0.76 to 9.58 for clinical failure. For radiographic failure, the sensitivity analysis showed RR 0.45, 95% CI 0.25 to 0.83 with a higher risk for radiographic failure in pulpectomy. The evidence was limited by the small number of trials included in the meta-analysis. Conclusions: Both pulpotomy and pulpectomy can be used successfully in the treatment of vital pulp exposure in primary incisors. Further high quality studies comparing between pulpotomy and pulpectomy in primary incisors with longer follow up period till exfoliation time are needed.


Introduction
Dental caries is an international public health challenge, especially among young children. Early childhood caries (ECC) is a serious public health problem in both developing and industrialized countries 1 . ECC has been considered to be at epidemic proportions in the developing world 2-6 . Treatment of ECC can be accomplished through different types of intervention. When caries extend to involve the pulp, various forms of pulp treatment have been used to treat and/or remove the pulp or to stimulate tooth repair. The choice of technique is as important as the choice between the different pharmacotherapeutic agents which are used in the treatment of primary teeth 7 .
Although pulpotomy is the treatment of choice for vital primary tooth pulp exposure throughout the pediatric dental literature 8 , the current trend amongst many dentists is to perform pulpectomies for the pulp treatment of carious vital primary anterior teeth 9 .
The most common materials used for pulpotomy are formocresol, ferric sulfate, also calcium hydroxide has been used, but with less long term success and more recently mineral trioxide aggregate (MTA) which is much more expensive 8,10 .
According to the latest recommendations of the American Academy of Pediatric Dentistry, both formocresol and MTA are strongly recommended to be used in pulpotomy with moderate quality of evidence. Ferric sulfate, laser and sodium hypochlorite are conditionally recommended but as for calcium hydroxide there is a recommendation against its use in pulpotomy 11 .
In pulpectomy, a resorbable material such as nonreinforced zinc/oxide eugenol (ZOE), a combination paste of iodoform and calcium hydroxide (Vitapex, Metapex) or a combination paste of zinc oxide and eugenol, iodoform and calcium hydroxide (Endoflas) are used to fill the canals 12 .
According to a Cochrane systematic review, there was no conclusive evidence supporting the superiority of one material for use in pulpectomy. Zinc oxide and eugenol, Metapex and Endoflas were found to be equally effective while with low quality of evidence zinc oxide and eugenol may be better than Vitapex but further research is required for confirmation 13 .
There are few studies that have compared pulpectomies with pulpotomies in vital primary incisors 14 . Therefore the claim that pulpotomies don't work in primary anterior teeth is not supported by high-quality evidence from research. Moreover two studies recently showed that there was no significant difference in success rates of pulpotomies and pulpectomies in the pulp treatment of asymptomatic vital primary incisors 9, 15 .
We therefore aimed to determine in patients with carious vital pulp exposure in primary incisors if pulpotomy is better than pulpectomy in terms of pain, soft tissue pathology, tooth mobility, pathological root resorption, periapical radiolucency, pulp canal obliteration and tooth survival based on evidence from randomized controlled trials.

Methods
Criteria for considering studies for this review Types of participants. Children with carious vital pulp exposure in primary incisors.

Types of interventions.
Pulpotomy and pulpectomy (root canal treatment) techniques with different medicaments.
Types of outcome measures Primary outcomes. We defined two primary outcomes: clinical failure and radiological failure.

Secondary outcomes.
According to the core set of component outcomes as specified by Smaïl-Faugeron et al. 16 these secondary outcomes were considered: Secondary clinical outcomes: pain, soft tissue pathology (gingival swelling, fistulous tract), pathological mobility and tooth survival.

Types of studies
Randomized controlled trials comparing pulpotomy and pulpectomy techniques in the treatment of carious vital pulp exposure in primary incisors were included.
Search methods Electronic search. We searched the electronic databases as the Cochrane library to 1/3/2018 and Pubmed to 1/3/2018. We developed detailed search strategies for each database searched. We placed no restrictions on the date of publication when searching the electronic databases. The search strategy included appropriate keywords, and Mesh terms when applicable; combined with Boolean operators "AND", "OR" and "NOT as shown in Table 1 and Table 2. We also searched OpenGrey for grey literature and ClinicalTrials.gov for ongoing trials.
Hand searching. We hand searched the reference lists of the included full text articles.

Data collection and analysis
Selection of studies. Two review authors (LG and AEB).independently scanned the titles of all reports identified by the search to determine whether the studies were relevant. They independently scanned selected abstracts to determine whether the study was relevant. After obtaining the full report for all relevant articles, they independently scanned the full reports and

Amendments from Version 2
In this version we edited the abstract to add the risk of bias tool and effect measures used. We corrected two lines in the search strategy and added the radiographic assessment in the Conclusion section. completed the data extraction form to determine whether the article should be included or excluded. Disagreements at each stage were resolved by discussion. If this did not resolve the disagreement a third author was invited to settle the disagreement. This did not occur over the course of this study.

REVISED
Data extraction and management. Two review authors (LG and AEB) collected the data independently using a specially designed data extraction form (Dataset 1 17 ).
For each trial, the following data were recorded: the year of publication, the country where the trial took place, detailed description of methodology, sample size, mean age of participants, duration of follow-up and reported outcomes. We contacted the authors of randomized controlled trials for missing information if needed.

Assessment of risk of bias in included studies.
Two review authors (LG and AEB) independently assessed the risk of bias in the included trials. The assessment was according to Cochrane risk of bias tool for quality assessment of randomized controlled trials described in the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0 (updated March 2011) 18 . Any disagreements between the two authors were resolved by discussion. If this did not resolve the disagreement a third author was invited to settle the disagreement. This did not occur over the course of this study.
Measures of treatment effect. Estimated effect size was calculated as risk ratios with 95% Confidence Interval (CI) for dichotomous outcomes. The unit of analysis was the tooth, because teeth were randomly assigned to intervention.
Data synthesis. The effect sizes and associated 95% confidence intervals were calculated using the Mantel-Haenszel method. If the results from trials were homogenous then fixed effect model was preferred. The statistical analysis was performed with the Review Manager program v.5.3 (RevMan) 19 .

Results
By searching the different databases, 14 references were identified after removal of the duplicates. By scanning the titles and abstracts, 4 studies were included as shown in the Prisma flow diagram Figure 1 and data extraction was performed (completed PRISMA checklist is available as Supplementary File 1).  Table 3. Ferric sulfate and mineral trioxide aggregate in pulpotomy was compared with zinc oxide and eugenol in pulpectomy in the Nguyen et al. trial 15 .

Pulpotomy
Two trials by Howley et al. 9 and Aminabadi et al. 20 used formocresol following hemostasis with a cotton pellet and zinc oxide and eugenol as capping material. One trial used ferric sulphate to achieve hemostasis and zinc oxide and eugenol as capping material, Casas et al. 21 , and one trial used ferric sulfate to achieve hemostasis, with mineral trioxide aggregate as a capping material, Nguyen et al. 15 .
Pulp access. Following caries removal and pulp exposure, the pulp chamber was accessed using a sterile no. 56 bur in a water-cooled high-speed handpiece and was refined using a sterile round bur in a slow-speed handpiece in three trials, Nguyen et al. 15 , Aminabadi et al. 20 and Casas et al. 21 .
In the trial by Howley et al. 9 the pulp chamber was unroofed using a no. 330 sterile bur in a water-cooled high-speed handpiece then the access was refined using a sterile round bur in a slowspeed handpiece.
Pulpotomy. The coronal pulp was amputated using a sharp spoon excavator in two trials by Howley et al. 9 and Aminabadi et al. 20 and was amputated using a sterile low-speed round bur in another two trials by Nguyen et al. 15 and Casas et al. 21 .

Root canal treatment.
In three trials Nguyen et al. 15 , Howley et al. 9 and Casas et al. 21 pulp tissue was removed en bloc using two or more endodontic files (Hedström files or K files), if the first attempt was unsuccessful, the procedure was repeated until all of the pulp tissue was removed.
In the fourth trial by Aminabadi et al. 20 an endodontic K file was introduced to the working length after a periapical radiograph was taken, and most of the pulp tissue was removed completely on the first attempt. If the first attempt was unsuccessful, the procedure was repeated and canals were generally enlarged three sizes past the initial file.   Table 4, Table 5, Table 6, and Table 7. The overall risk of bias was low in two trials, Nguyen et al. 15 and Aminabadi et al. 20 . The risk of bias was unclear for clinical assessment in two trials, Howley et al. 9 and Casas et al. 21 , and also for random sequence generation in Casas et al. 21 . One trial, Casas et al. 21 , showed high risk of bias due to high percentage of dropped out cases. The risk of bias of included studies is also summarized in Figure 2.

Synthesis of results.
Only three trials were included in the meta-analysis by Nguyen et al. 15

Clinical failure
At the longest follow up period the pooled results showed no statistically significant difference in clinical failure for pulpotomy compared to pulpectomy (RR 2.69 ,95% CI 0.76 to 9.58) as shown in Figure 3.
The Howley et al. study was excluded from the meta-analysis of clinical failure as there were no events in both groups.
It is the standard practice for the meta-analysis for risk ratio when there are no events in both arms 18 . Quote: The radiographs were evaluated independently by 2 standardized and calibrated examiners who were not otherwise involved in the study. It is not possible to blind the assessors due to the nature of treatment received.
Incomplete outcome data (attrition bias) Low risk 29 study patients, 3 patients failed to return for follow-up. Random sequence generation (selection bias) Low risk Computer-generated simple random numbers sequence with a one to one allocation ratio.

Allocation concealment (selection bias) Low risk
Quote: Allocation occurred after induction of general anesthesia to ensure allocation concealment. The pediatric dentist, nurse, or assistant directed subjects at the time of dental surgery to the appropriate treatment group they had been assigned to by the investigator.
Blinding of participants and personnel (performance bias) Low risk Quote: All other contributors were blinded to generation and implementation of the treatment assignment Blinding of clinical outcomes assessment Low risk Quote: A single investigator, who did not perform any pulp therapy or participate in radiographic evaluation, performed all clinical assessments.
Blinding of radiological outcomes assessment Low risk Quote: Two experienced pediatric dentists who did not participate in protocol development or treatment performed radiographic assessments. It is not possible to blind the assessors due to the nature of treatment received.

Reporting bias Low risk
We revised the protocol that was registered with the ClinicalTrials.gov Protocol and Registration System (ID no. NCT02019563)

Radiographic failure
The pooled results showed RR 0.79, 95% CI 0.46 to 1.21 in radiographical failure for pulpotomy compared to pulpectomy as shown in Figure 4 but considerable heterogeneity was presented in the meta-analysis of radiographic failure due to the presence of one outlying study of Aminabadi et al. so we performed the analysis both with and without this study as part of a sensitivity analysis as shown in Table 8. Quote: Quality assurance checks were performed by 1 of the investigators (MAL), who did not provide treatment or review postoperative radiographs, to ensure that the investigators who provided treatment complied with the randomization protocol Blinding of participants and personnel (performance bias) All outcomes Low risk It is not possible to blind the operator and the participant blinding is ineffective on outcomes Blinding of clinical outcomes assessment unclear risk Insufficient information to make a clear judgement.
Blinding of radiological outcomes assessment Low risk Quote: Two independent pediatric dentists who were not otherwise involved in the investigation evaluated the radiographs. It is not possible to blind the assessors due to the nature of treatment received.
Incomplete outcome data (attrition bias) All outcomes

High risk
Quote: Of the enrolled participants, 64% returned for at least 1 evaluation. 36% drop out in the pulpotomy group and 48% drop out in the root canal treatment group at 2 years follow up Quote: Two clinicians who did not perform any treatments analyzed the clinical and radiographic outcomes It is not possible to blind the assessors due to the nature of treatment received.
Incomplete outcome data (attrition bias) Low risk 4 subjects were dropped out of 50 patients.
The sensitivity analysis for radiographic failure showed RR 0.45, 95% CI 0.25 to 0.83 with statistical significant difference between pulpotomy and pulpectomy and a higher risk for radiographic failure in pulpectomy

Soft tissue pathology
The pooled results showed no statistically significant difference in soft tissue pathology for pulpotomy compared to pulpectomy (RR 3.11, 95% CI 0.54 to17.7)

Pathological radiolucency
The results of two trials showed considerable statistical heterogeneity so no meta-analysis were performed for this outcome.

Pathological root resorption
The pooled results of two trials showed no statistically significant difference in pathologic resorption for pulpotomy compared to pulpectomy (RR 1.5, 95% CI 0.56 to 4.04). Summary data using the study data extraction form

Discussion
Pulp therapy is performed to preserve primary teeth and maintain its developmental, esthetic, and functional capabilities. Pulpotomy and root canal therapy have been both performed as techniques for the management of asymptomatic vital primary incisors with large carious lesions where removal of caries will lead to pulp exposure 20 . However the preference of many pediatric dentists to perform pulpectomy in primary incisors was due to the fact that they were taught to do so in their pediatric dentistry residencies and not due to evidence from high quality research 9 . The aim of this systematic review was to compare between pulpotomy and pulpectomy clinically and radiographically in the treatment of carious vital pulp exposure in primary incisors.
Upon performing our systematic search there were only four randomized controlled trials that have compared pulpotomy and pulpectomy outcomes in vital primary incisors 9,15,20,21 . After exclusion of one trial due to its high risk of bias, only three trials were left to be included in the meta-analysis. The data of the longest follow up period was included as the follow ups were close to each other ranging from 15 months to 24 months and they best reveal the efficacy of the performed techniques.
The results were calculated with risk ratio (RR) effect measure and confidence intervals (CIs). The pooled results of the clinical failure outcome showed the relative risk RR was 2.69 with 95% CI from 0.76 to 9.58, the CI including the number 1 means that there was no statistical significant difference between pulpotomy and pulpectomy cases.
For radiographic failure the RR was 0.79 with 95% CI from 0.25 to 2.42 but considerable statistical heterogeneity was detected due to the presence of one outlying study of Aminabadi et al. with results that conflict the other studies so sensitivity analysis was performed that showed that RR was 0.45 with 95% CI 0.25 to 0.83 with higher risk of radiographic failure for pulpectomy rather than pulpotomy.
Although there is no clinical diversity among the included trials that may have led to this statistical heterogeneity, we may relate it to that the criteria of radiographic assessment in Aminabadi et al, study was not clearly specified.
The exclusion of a study from a meta-analysis based on their result may introduce bias so the results must be interpreted with an appropriate degree of caution and futher investigaton is required 18 . For the clinical outcomes pain and soft tissue pathology, the pooled results for these outcomes showed no statistically significant difference between pulpotomy and pulpectomy while pathologic mobility was only reported for one incisor in one trial.
The radiographic outcomes included periapical radiolucency and pathologic root resorption. For pathologic resorption, the pooled results showed no statistically significant difference between pulpotomy and pulpectomy. We considered the tooth to be scored with pathologic root resorption if it showed perforating internal root resorption or large external root resportion while those teeth showing contained internal root resorption or questionable external root resorption were not counted. For periapical radiolucency, only frank radiolucencies were counted and not questionable ones but the results of the trials included was inconsistent.
Pediatric dentists do not consider the radiographic pathological changes as questionable radiolucencies or limited pathological root resorptions to be an absolute indication for extraction taking into consideration the absence of clinical pathological signs or symptoms 15 .
Overall failure was reported for two pulpotomized incisors in two trials. On the other hand tooth survival although it is an important outcome, but it is not commonly reported.
This study showed that there is no statistical significant difference in clinical success rates of pulpotomy and pulpectomy with different medicaments in the treatment of carious vital pulp exposure in primary incisors while radiographically, pulpectomy may have a higher risk for radiographic failure than pulpotomy and this refutes the misconception among some pediatric dentists that pulpotomy does not work in primary incisors.
The evidence was limited by the small number of trials included in the meta-analysis. The overall risk of bias of primary studies was low for three trials, except for the unclear risk for blinding of clinical assessment which was not effective., We did not have access to all the trial protocols to assess the selective reporting bias except for only one trial.

Conclusion
There was no statistical significant difference in the clinical success rates of pulpotomies and pulpectomies in the pulp treatment of carious vital pulp exposure in primary incisors while radiographically, pulpectomy may have a higher risk for radiographic failure than pulpotomy. We recommend teaching of pulpotomy as a treatment option for vital pulp exposure in primary incisors in pediatric dentistry residency programs and further high quality studies comparing between pulpotomy and pulpectomy in primary incisors with a longer follow up period till exfoliation time.

Grant information
The author(s) declared that no grants were involved in supporting this work.

Author information
LG Reviewer Expertise: systematic reviews, pediatric dentistry, dental materials, dental trauma, carie I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Author Response 20 Jun 2019 , National Research Centre, Egypt, Nasr city, Egypt Lamia Gadallah Dear Dr. Marcela, We are honored by your reviewing to our article and we took into consideration all your valuable corrections and comments. Firstly, in the abstract section: The risk of bias tool and meta analysis method used was added. The relative risk is called the risk ratio and both terms are mentioned in Cochrane Handbook for systematic reviews of interventions. Radiographically the risk ratio value was 0.45 95%CI 0.25 to 0.83. These values are less than one which means higher risk of radiographic failure in the control(pulpectomy) group. Secondly, in the methodology section: The authors chose the primary outcomes according to a clinical point of view that there is a set of outcomes clinically and radiographically that collectively indicate the success or failure of the technique and you are absolutely right that those secondary outcomes are the single outcomes that composed the composite primary outcome, and we preferred to show their effect separately as secondary ones due to their less importance as single outcomes. Yes the search was last updated in march 2018. In the search strategy, we thank you for the two mistakes you have highlighted. By revising the saved history of search strategy, we found two editing mistakes, first line #22 is composed of #19,#20 #21 and not #18. Line 24 is combined with AND not OR because we wanted the studies of incisors and pulpotomy with incisors and pulpectomy, that's how the actual search was.performed, the mistakes were during editing tables. The I2 is considered homogenous up till 50%. Thirdly, the results section: Figure 2 is according to Risk of bias summary figure in Cochrane Handbook where green represents low risk, red represents high risk and blank is unclear risk. The Howley et al study was automatically excluded from the forest plot as there are zero events in both arms, we left it only in the table to show that zero events are not estimable In the periapical radiolucency outcome the two trials included were Howley et al. and Aminabadi et al, , the study of Nguyen et al. was excluded due to lack of values in this outcome, only odds ratio was reported with no prevalence percentage in the control or non exposure group so we could not convert the odds ratio to risk ratio. Lastly in the conclusion section: Radiographic assessment was added. The systematic review and meta-analysis regarding pulpotomy versus pulpectomy in the treatment of vital pulp exposure in primary incisors is well designed and that the methodology and discussion are well written. One question I have is: was the protocol of this review registered before carrying out the research? Some mistakes in writing the manuscript are present which are:

No competing interests
Please be sure the reference number is written right after the author name through-out the whole manuscript.
In the background section of the abstract: This study aimed to assess the (effectiveness) of pulpotomy and pulpecomy (pulpectomy) in treatment of carious vital pulp exposure in primary incisors.
In the conclusion section of the abstract; Further high quality studies comparing between pulpotomy and pulpectomy in primary incisors with longer follow up period till exfoliation time (are needed). Page 2, introduction section: In pulpectomy(,) a resorbable material such as nonreinforced According to a Cochrane systematic review(,) there was no Endoflas were found to be equally effective(.) (W)hile(,) with (a) low quality of evidence(,) zinc oxide and eugenol may be better than We(,) therefore(,) aimed to determine in patients with carious vital pulp exposure in primary incisors(,) if pulpotomy is better than