Recurrence rate following treatment for primary multicystic ameloblastoma: systematic review and meta-analysis


Opinions regarding the treatment of multicystic ameloblastoma are divergent due to its benign nature and the high rate of recurrence if not adequately excised. The aim of the present study was to perform a systematic review of the literature for a qualitative and quantitative assessment of studies addressing primary multicystic ameloblastoma with regard to treatment and recurrence. Searches were conducted of the Ovid Medline and Embase databases for articles published up to January 2014. Based on predefined eligibility criteria, studies were selected in a two-stage screening process conducted by two independent reviewers. Quality assessment of the selected articles was performed using the modified criteria of the Agency for Healthcare Research and Quality. The meta-analysis was performed using Review Manager (RevMan) software. Statistical heterogeneity was investigated by performing a χ 2 test at the 5% significance level ( P < 0.05) and determining I 2 . The relative risk of recurrence was 3.15-fold greater (95% confidence interval 1.98–5.00) when conservative treatment was performed on primary multicystic ameloblastoma in comparison to radical treatment ( P < 0.00001 for treatment effect; I 2 = 0% and P = 0.48 for heterogeneity). The findings justify the treatment of primary multicystic ameloblastoma with bone resection.

Ameloblastoma was originally described by Cuzack in 1827 and subsequently detailed by Broca in 1868. The World Health Organization (WHO) classifies ameloblastoma as a tumour formed by odontogenic epithelium with mature fibrous stroma without an odontogenic ectomesenchyme. Ameloblastoma accounts for 1% of tumours that affect the oral–maxillofacial complex and 13–58% of odontogenic tumours, and is classified into four types based on clinical, histological, and radiographic findings: unicystic, multicystic, peripheral, and desmoplastic. Each type has a distinct clinical behaviour and requires different forms of treatment.

Multicystic ameloblastoma is considered the most common variant and consists of a locally aggressive tumour that mainly affects the body and posterior portion of the mandible in patients in the third and fourth decades of life ; there is no predilection for the male or female sex. The diagnosis is generally made through a combination of imaging examinations (X-ray and computed tomography) and biopsy to determine the histological type. In an X-ray, the tumour exhibits a well-defined unilocular or multilocular osteolytic image surrounded by a radiopaque margin, often with expansion of the bone cortex, with the possibility of root resorption and occasional association with an impacted tooth.

Treatment is classified as conservative or radical. Conservative treatment includes enucleation, curettage, and surgical excision with peripheral osteotomy or other adjuvant therapy, such as cryotherapy or Carnoy’s solution. Radical treatment consists of bone resection. In the mandible, resection can be completed through segmental osteotomy or a mandibulectomy, or can be marginal (when the lower border is preserved, thereby maintaining bone continuity). In the maxilla, radical treatment comprises maxillectomy, which is classified as partial or total based on the extent of the resection.

Surgical treatment of the multicystic variant remains a controversial issue, as the tumour has a high rate of recurrence if not adequately excised, but the metastatic potential is low. With conservative treatment, 55–90% of cases recur, whereas the recurrence rate with more radical treatment is 15–25%; however, in the latter case, the patient experiences serious aesthetic and functional impairment as well as the need for reconstructive surgery. Nakamura et al. found recurrence rates of 7.1% following radical surgery and 33.3% following conservative surgery, which are much lower rates than those reported in other studies, encouraging the use of more conservative treatment. Similarly, Hasegawa et al. reported a recurrence rate of 43.5% following conservative treatment.

Hong et al. concluded that treatment tends to be radical. When age and location are the same, going from conservative treatment to resection with a bone margin, or going from the latter to segmental resection or maxillectomy, the recurrence rate is diminished by 20%. The authors also stated that tumours with a multicystic radiographic appearance have a 3.02-fold greater chance of recurring in comparison to those with a unicystic appearance and that patients with only bone expansion without rupture of the cortex have a lower chance of recurrence in comparison to those with rupture of the bone cortex.

Due to its benign nature and slow growth, many surgeons believe that the treatment of ameloblastoma should initially be conservative, with radical surgery only performed when necessary after a recurrence. In other words, the diagnosis of a multicystic ameloblastoma should not necessarily indicate immediate bone resection, and treatment should be performed with the complete removal of the tumour while preserving the lower portion of the mandible whenever possible.

Considering the divergent opinions on the treatment of multicystic ameloblastoma, the aim of the present study was to perform a systematic review of the literature for a qualitative and quantitative assessment of studies addressing primary multicystic ameloblastoma with regard to treatment and recurrence.

Materials and methods

Focused question

The following question guided this study: What form of treatment for primary multicystic ameloblastoma results in the lowest rate of recurrence?

Search strategy

Searches were performed of the Ovid Medline and Embase databases for articles published up to January 2014. The search strategy involved combinations of medical subject heading (MeSH) terms and key words for the Ovid Medline database and the appropriate modifications for the Embase database. The search strategy was as follows: (1) Population: “Ameloblastoma” OR “Jaw neoplasms” OR “Mandibular disease” OR “Maxillary disease” OR “Ameloblastoma$” OR “{(jaw or mandibular or maxillary) and neoplasm$}” OR “{(jaw or mandibular or maxillary) and disease$}” OR “{(jaw or mandibular or maxillary) and tumo$}” OR “(odontogenic and tumo$)”. (2) Intervention: “Curettage” OR “Decompression, surgical” OR “Decompression” OR “Chloroform” OR “Ethanol” OR “Acetic acid” OR “Cryosurgery” OR “Cryotherapy” OR “Curettage$” OR “Decompression$” OR “{(carnoy or carnoy’s) and solution$}” OR “Cryotherap$” OR “Cryosurg$” OR “Enucleation$” OR “{(mandibular or jaw or maxillary or marginal or box or wide or radical or conservative) and resection$}” OR “Mandibulectomy” OR “Maxillectomy” OR “(peripheral and osteotomy)” OR “(liquid and nitrogen)” OR “{surgical and (treatment or management)}” OR “{(radical or conservative) and surger$}” OR “Excision$” OR “Marsupialization$”. (3) Outcomes: “Neoplasm recurrence, local” OR “Retreatment” OR “Treatment outcome” OR “Morbidity” OR “Recur$” OR “Relapse$” OR “(long and term and effect$)” OR “Morbidit$” OR “Retreatment$” OR “(effective and therap$)” OR “(Treatment$ and outcome)” OR “(Follow and up)”.

The key words were then combined in the following way: Population AND Intervention AND Outcomes.

Manual searches of the reference lists of the articles were also performed to find further studies that were not included in the electronic databases.

Outcome measure

The primary outcome of the study was the recurrence rate.

Eligibility criteria

The inclusion criteria were articles addressing the treatment of primary multicystic ameloblastoma and level of evidence up to the fourth level, as established by the Oxford Centre for Evidence-Based Medicine. The exclusion criteria were treatment for ameloblastomas in uncommon locations, articles that referred only to the desmoplastic histological type, articles that referred only to one type of treatment, articles that referred only to the treatment of giant ameloblastomas, cases series with fewer than 10 cases, mean postoperative follow-up period of less than 5 years or not specified, treatment only for individuals less than 20 years of age (due to the tendency towards conservative treatment in this age group), use of histological classification different from that recommended by the WHO, failure to describe the histopathological analysis, failure to distinguish unicystic and multicystic tumours from the histological standpoint regarding the treatment employed, failure to specify treatment, failure to describe recurrences, and study with the same population as that of another study included.

Screening methods

The article selection process was conducted in two steps by two independent researchers (R.A.C.A. and J.C.B.) who were initially blinded to each other’s results. After a comparison of the findings, disagreements regarding the inclusion or exclusion of a given article were resolved by consensus. If necessary, a third researcher (E.S.S.A.) was consulted to decide whether or not a study should be included. In the first step, titles and abstracts were read for the elimination of irrelevant articles and those that did not meet the eligibility criteria. The second step involved the full-text analysis of the studies selected independently by the two researchers based on the eligibility criteria. Quality assessment and data extraction were performed only for the articles selected for inclusion in the present review. Agreement between the two researchers was calculated using the kappa statistic in both steps of the selection process.

Data extraction

Data extraction was performed using a chart designed specifically for the present review addressing the type of study, description of the study population, demographic aspects of the sample, similarity of the study groups, description of the histopathological analysis, adequate description of the treatment, description of the postoperative follow-up time, and description of recurrences following primary treatment.

Methodological quality assessment

The risks of bias of the studies selected was assessed independently by the two researchers (R.A.C.A. and J.C.B.) using the modified scale of the Agency for Healthcare Research and Quality for observational studies. This scale has nine evaluation criteria: study question, study population, comparability of subjects for all observational studies, exposure or intervention, outcome measures, statistical analysis, results, discussion, and funding or sponsorship. The scale offers an overall score as well as scores for the evaluation of the individual components, which are considered ‘adequate’ (A) when the study clearly meets the specific item, ‘inadequate’ (I) when the study does not clearly meet the specific item, or ‘not reported’ (NR) when the authors failed to cite a specific item.

Statistical methods

A meta-analysis was performed for the quantitative evaluation of the data extracted using Review Manager software (RevMan version 5.3). The 95% confidence interval (CI) was calculated. Results with a P -value of <0.05 were considered indicative of statistical significance. A forest plot was drawn to enable the evaluation of the findings in a global and individual fashion for each article. Heterogeneity among the studies was calculated using the χ 2 test and the determination of I 2 (values ≥50% indicate substantial heterogeneity). The meta-analysis of random effects was performed in cases of statistical evidence of heterogeneity. Otherwise, an analysis of fixed effects was performed. The relative risk (RR) was used to determine the effect size for the comparison of the incidence of recurrence between conservative treatment and radical treatment, and the comparison of the incidence of recurrence between marginal resection and segmental resection. The studies were combined to furnish a total estimate with a 95% CI.


Search sequence and quality assessment of selected publications

The search strategy yielded a total of 11,790 articles: 6744 from Embase and 5046 from Ovid Medline. The articles were downloaded to the Mendeley program, which identified 548 duplicates, leaving 11,242 articles with which to begin the selection process. After reading the titles and abstracts, 71 articles were selected for full-text analysis. Eight articles met the eligibility criteria, but only seven were submitted to the evaluation of the risk of bias and data extraction, since two articles were by the same authors and used the same sample, the first of which was considered for analysis ( Table 1 ).

Table 1
Risk of bias among studies analyzed.
Domains Elements Chapelle et al., 2004 Fregnani et al., 2010 Hasegawa et al., 2013 Hertog et al., 2012 Hong et al., 2007 Junqueira, 2003 Nakamura et al., 2002
Study question Clearly focused and appropriate question A A A A A A A
Study population Description of study populations A A A A A A A
Sample size justification NR A A A A A A
Comparability of subjects Specific inclusion/exclusion criteria I A I A I I A
Criteria applied equally to all groups a a a a a a a
Comparability of groups at baseline with regard to disease status and prognostic factors a a a a a a a
Study groups comparable to non-participants with regard to confounding factors a a a a a a a
Use of concurrent controls a a a a a a a
Comparability of follow-up among groups a a a a a a a
Exposure or intervention Clear definition of exposure A A A A A A A
Measurement method standard, valid, and reliable A A A A A A A
Exposure measured equally in all study groups A A A A A A A
Outcome measurement Primary/secondary outcomes clearly defined A A A A A A A
Outcomes assessed blind to exposure or intervention status a a a a a a a
Method of outcome assessment standard, valid, and reliable NR A A A A A A
Length of follow-up adequate for question A A A A A A A
Statistical analysis Statistical tests appropriate NR A A A A A I
Multiple comparisons taken into consideration NR A A A A A I
Modelling and multivariate techniques appropriate NR A NR A A A NR
Power calculation provided NR A NR NR A NR NR
Assessment of confounding variables a a a a a a a
Dose–response assessment, if appropriate a a a a a a a
Results Measure of effect for outcomes and appropriate measure of precision NR A A A A A A
Adequacy of follow-up for each study group A A A A A A A
Discussion Conclusions supported by results with biases and limitations taken into consideration A A A A A A A
Funding or sponsorship Type and sources of support for study NR A A NR A NR NR
A, adequate; I, inadequate; NR, not reported.

a Not applicable to the study design.

Whenever necessary, the authors were contacted for clarifications and additional information needed for the evaluation of the article. Figure 1 shows the flowchart of the selection process. A table was created for the 64 excluded articles and the reasons for their exclusion after the full-text analysis ( Table 2 ). The level of agreement between the two researchers was calculated using the kappa coefficient; κ = 0.82 for the reading of the titles and abstracts and κ = 0.88 after the full-text analyses, demonstrating excellent agreement.

Fig. 1
Flowchart of the article selection process.

Table 2
Articles excluded and the reasons for their exclusion.
Articles excluded Reason for exclusion
Adekeye Failure to describe the histopathological analysis
Adekeye and Lavery
Arotiba et al.
Becelli et al.
Chana et al.
Chukwuneke et al.
Dandriyal et al.
Daramola et al.
Hammarfjord et al.
Jeblaoui et al.
Jackson et al.
Olaitan et al.
Olaitan et al.
Adebayo et al. Failure to distinguish unicystic and multicystic tumours from a histological standpoint regarding treatment employed
Bataille et al.
Eckardt et al.
Franca et al.
Hatada et al.
Pilz and Nitzschke
Pinsolle et al.
Rastogi et al.
Ruhin-Poncet et al.
Tanaka et al.
Tsaknis and Nelson
Vedtofte et al.
Carlson and Marx Failure to specify treatment
Crezoit et al.
Kim and Jang
MacDonald-Jankowski et al.
Reichart et al.
Sampson and Pogrel
Wu and Chan
Wu et al.
Ueno et al. Histological classification different from WHO
Ueno et al.
Progrel and Montes Not up to the fourth level of evidence based on the Oxford Centre for Evidence-Based Medicine
Krishnapillai and Angadi Follow-up of <5 years or not specified
Ghandhi et al.
Albuquerque et al.
Arotiba et al.
Bianchi et al.
Chaine et al.
Olusanya et al.
Ord et al.
Rapidis et al.
Siar et al.
Singappulige et al.
Vayvada et al.
Bataineh Only one form of treatment employed
Curi et al.
Nastri et al.
Adebayo et al. Treatment of children and adolescents only
Butt et al.
Chidzonga et al.
Carneiro et al. Case report (fewer than 10 cases)
Valcarce Gomez et al.
Schafer et al. Treatment of ameloblastoma in an uncommon location
Vongsa et al. Failure to describe recurrences
Hertog et al. Treatment was not for a primary tumour
Hertog and van der Waal Study with the same population as another study included
WHO, World Health Organization.

Study characteristics

The seven articles selected for data extraction involved a total of 394 cases of multicystic/solid ameloblastoma undergoing primary surgical treatment and with a postoperative follow-up period of at least 5 years. Table 3 displays the median age of the patients, number of multicystic ameloblastomas, the treatments performed, number of recurrences, follow-up time, and recurrence by histological type.

Table 3
Median age of patients, distribution of treatments, recurrence, and postoperative follow-up periods in the studies analyzed.
Authors and year of publication Median age (years) Number of multicystic ameloblastomas Treatment Recurrence Mean postoperative follow-up (years) Recurrence by histological type
Chapelle et al., 2004 42.5 14 Enucleation + Carnoy’s solution n = 4 1 9.3 2 follicular
Enucleation n = 6 3 2 follicular/plexiform
Marginal resection n = 2 0
Segmental resection n = 2 0
Fregnani et al., 2010 33.2 113 Curettage n = 19 3 9.7 12 follicular
Curettage + cryotherapy n = 47 14 8 plexiform
Segmental resection n = 47 8 3 acanthomatous
1 desmoplastic
1 hemangiomatous
Hasegawa et al., 2013 38.8 29 Enucleation n = 10 4 5 6 follicular
Marsupialization/enucleation n = 6 4 3 plexiform
Enucleation + curettage n = 7 2 1 desmoplastic
Marginal resection n = 3 0
Segmental resection n = 3 0
Hertog et al., 2012 31.6 28 Enucleation n = 22 14 8.3 7 follicular
Segmental resection n = 6 0 4 plexiform
3 follicular/plexiform
Hong et al., 2007 34.5 147 Enucleation n = 97 40 8 32 follicular
Marginal resection n = 31 4 9 acanthomatous
Segmental resection n = 19 1 4 granular
Junqueira, 2003 44.5 12 Enucleation + curettage n = 5 3 5 2 follicular
Marginal resection n = 2 1 1 plexiform
Segmental resection n = 5 1 1 acanthomatous
1 granular
Nakamura et al., 2002 34.1 51 Marsupialization + enucleation n = 11 8 5 5 follicular
Enucleation + curettage n = 11 2 5 plexiform
Marginal resection n = 4 0 3 follicular/plexiform
Segmental resection n = 25 3
Total 37.0 394 116 7.2
Only gold members can continue reading. Log In or Register to continue

Jan 16, 2018 | Posted by in Oral and Maxillofacial Surgery | Comments Off on Recurrence rate following treatment for primary multicystic ameloblastoma: systematic review and meta-analysis
Premium Wordpress Themes by UFO Themes