Comparison of survival outcomes between ameloblastic carcinoma and metastasizing ameloblastoma: A systematic review

Abstract Purpose To investigate and compare the demographic data, occurrence of recurrence and metastasis, and survival prognosis between ameloblastic carcinoma (AC) and metastasizing ameloblastoma (MA), based on appropriate and currently accepted eligible diagnostic criteria, in a systematic review of the literature. Methods An electronic search was undertaken, last updated in December 2021. Eligibility criteria included publications having enough clinicopathological information to confirm the diagnosis of these tumors. Results Seventy‐seven publications reporting 85 ACs and 43 MAs were included. Both tumors were more frequent in mandible and showed different clinical profiles regarding patients' sex and age. There was no difference in the estimated cumulative survival between patients diagnosed with these tumors. Metastases mainly affected the lungs, followed by cervical lymph nodes. The mean time between the first metastasis and the last follow‐up was higher for MA (p = 0.021). In addition, MA patients remained alive longer than AC patients after the first metastasis diagnosis (p = 0.041). Considering only the cases that metastasized, a higher ratio of AC patients died in comparison to MA patients (p = 0.003). The occurrence of recurrence was associated with a conservative primary treatment with both AC (p < 0.001) and MA tumors (p = 0.017). Multiple recurrent events were associated with conservative primary therapies with MA (p < 0.001) but not with AC (p = 0.121). Conclusion In addition to some demographic differences, ACs that metastasize present a worse prognosis than MA. As conservative procedures are associated with multiple recurrent events, this treatment modality should be avoided for both tumors.

authors as being either AC or MA, even not having these terms in the title of the article, were also re-evaluated by the present study's authors. Moreover, the publication needed to have a histological image and/or a minutely and consistent histopathological description of the metastatic lesion.

| Inclusion criteria
Concerning the histopathology, both primary and metastatic lesions needed to have histological features of benign ameloblastoma. 1

| Definitions
Treatments were classified either as conservative or aggressive/radical. Aggressive management was defined as any treatment in which the tumor is neither violated nor directly manipulated. 5 Either marginal resection or resection with a continuity defect fall within this perimeter.
Other types of treatment were considered as conservative, which included marsupialization, curettage, enucleation, debulking, excision, and chemotherapy and/or radiotherapy alone.
Adjunctive therapies were defined as another treatment used together with the primary treatment, and could comprise neck resection, chemotherapy, radiotherapy, or a combination of these.

| Exclusion criteria
Exclusion criteria were immunohistochemical studies, histomorphometric studies, radiological studies, genetic expression studies, histopathological studies, cytological studies, cell proliferation/apoptosis studies, in vitro studies, and review papers, unless any of these publication categories had reported any cases that fulfill the inclusion criteria. Moreover: • Insufficient clinicopathological data to describe each individual case; • Histopathological images that did not allow the diagnosis to be confirmed; • Histopathological images that demonstrated only an increase in cellularity, not being sufficient to suggest the diagnosis of AC; • Only description of an alleged MA, without the microscopic histopathological documentation and/or detailed description of the metastatic lesion.

| Study selection
The titles and abstracts of all reports identified through the electronic searches were read independently by the authors. For studies appearing to meet the inclusion criteria or for which there were insufficient data in the title and abstract to make a clear decision, the full report was obtained. Disagreements were resolved by discussion between the authors. The clinicopathological features of the tumors reported by the publications were thoroughly assessed by five authors of the present study, to confirm the tumors' diagnosis.
RefWorks Reference Management Software (Ex Libris, Jerusalem, Israel) was used in order to detect duplicate references in different electronic databases.

| Data extraction
The review authors independently extracted data using specially designed data extraction forms. Any disagreements were resolved by discussion. For each of the identified studies included, the following data were then extracted on a standard form, when available: patient's sex, patient's age at the primary lesion, duration of the lesion previously to treatment, location of the primary lesion (maxilla/mandible), occurrence and location of metastatic lesions, treatment performed (surgery, chemotherapy, radiotherapy, neck dissection), recurrences, death, the time between primary lesion and metastasis, and follow-up.
Contact with authors for possible missing data was performed.   Table 1 compares the demographic and clinical features between the cases of AC and MA. Both tumors were more prevalent in the mandible than in the maxilla. AC was more common in men than in women, whereas MA was slightly more frequent in men. The mean age for the primary lesion was higher for AC than for MA (p < 0.001). About 70% of the AC cases (when the information was available) were de novo.

| Description of the studies and analyses
Many patients were submitted to a plethora of surgical approaches for the removal of the tumor, with or without adjunctive therapies, with diverse sequences of procedures. For example, radiotherapy first followed by surgery, then chemotherapy + radiotherapy, or many surgeries after multiple recurrences. Therefore, the influence of the first treatment on prognostic variables (recurrence, metastasis, and death) was evaluated. Information on primary therapy was available for 61 (71.7%) cases of AC and 40 (93%) cases of AM. Ten categories of procedures were performed at the initial treatment. For analytical purposes, we subdivided these treatments into two subgroups: conservative primary therapy and aggressive primary therapy (Table 2). An aggressive primary therapy was performed for 43 of 61 (70.5%) ACs, while 21 of 40 (52.5%) MAs received a more radical approach at initial management ( Table 2).
T A B L E 1 Detailed description of ameloblastic carcinoma (AC) and malignant ameloblastoma (MA) included in the study (information was not always available for all variables for all cases)

| DISCUSSION
The demographic data of AC and MA show that both tumors were more prevalent in the mandible and AC was more common in men than in women, whereas MA was slightly more common in men. The mean age for the primary in AC (49.6 ± 14.9 years) was higher than for AC (31.7 ± 6.6 years). Despite the limitations of the study, this data gives support to the assumption that they are distinct clinicpathologic entities.
According to the results of the present study, the difference of  The combination of several therapies impaired a proper analysis of the efficiency of different treatments. However, considering that the initial management is mandatory for a better prognosis of any disease, the primary therapy effect on some prognostic factors was analyzed. Interestingly, a higher proportion of patients with MA (47.5%) received initial conservative management than AC patients (29.5%).
Conservative primary procedures were associated with the occurrence of recurrence in both AC and MA tumors, and multiple recurrent events with MA. These data highlight the value of assertive decision-making for managing ameloblastoma patients. Conservative procedures such as those described in Table 2 should be avoided.
As the authors of the present review established that one of the criteria for the identification of a MA was a metastasis of histologically well-differentiated ameloblastoma, the diagnosis of MA based on a literature review was not a simple task, due to the lack of documentation of both primary and metastatic tumors in many publications.
Almost 20 years ago, Reichart  The results of our study have to be interpreted with caution because of its limitations. First, all included studies were retrospective reports, which inherently result in errors, with incomplete records. It was not possible to retrieve information on all variables from all cases, which would have improved the quality of the statistical analyses. 10 Moreover, the authors of this study needed to rely on the printed histopathological exams of the metastatic tumor to select valid cases of MA, rather than having access to the original pathology slides. Second, many of the published cases had a short follow-up, which could have led to an underestimation of the actual survival rate.

| CONCLUSION
In addition to some demographic differences, patients with AC that metastasize present a worse prognosis than patients with MA. As conservative procedures are associated with multiple recurrent events, this treatment modality should be avoided for both tumors.