Introduction

Ameloblastoma of the jaws is the most common epithelial odontogenic tumor. It is sub-classified into intra-bony conventional and unicystic varieties. The unicystic ameloblastoma comprises of luminal, intraluminal and mural subtypes [1, 2]. Ameloblastomas have no cytological atypia, are defined as benign even though they may display locally aggressive biological behavior [1, 2]. The current standard of care mandates bone resection with a 1–2 cm margin of safety beyond the radiological borders for both the conventional and unicystic mural subtype, due to similar risks of recurrence [2]. The resulting bony defect frequently calls for immediate reconstruction, which is usually based upon a vascularized free flap harvested from an additional surgical site. These extensive surgical procedures might permanently affect jaw and tooth development, function, and esthetics, and inevitably have a negative psychological impact, especially among young patients [3]. Furthermore, such extensive surgery bears morbidity that might require additional surgical procedures. These sequelae mandate the search for enhanced treatment approaches.

Identification of the BRAFV600E driver mutation in mandibular ameloblastomas as well as other mutations usually associated with maxillary ameloblastomas almost a decade ago has led to the search for targeted therapy for ameloblastomas of the jaws [4,5,6].

Reported attempts to treat ameloblastomas with BRAF inhibitors (BRAFi) [with/without mitogen-activated protein kinase (MEK) inhibitors (MEKi)] as a last line treatment, have been successful for seven patients with recurrent inoperable conventional ameloblastomas, even some with lung metastatic deposits [7,8,9,10,11,12,13,14]. BRAF-targeted therapy was used effectively for the first time in a neoadjuvant protocol in a small series of young patients with unicystic, mural type ameloblastoma [15]. Tumors responded with a significant reduction in size to a point that enabled conservative removal of the residual tumor, enabling jaw preservation and ad integrum mandibular regeneration. These young patients were part of a larger series of 11 patients (eight unicystic mural type, three conventional ameloblastomas), successfully treated with neoadjuvant BRAFi (with/without MEKi) [16]. An additional case of ameloblastoma with a short-term neoadjuvant BRAFi protocol has also been reported [17].

The effect of BRAF mutations on proliferation and survival of affected cells is the most recognized and widely investigated outcome. However, some of the multifaceted additional impacts that this genetic aberration has upon signaling pathways in both the mutated cells themselves as well as on the surrounding, non-mutated cells within the tumor microenvironment are important as well (Fig. 1A). Oncogenic-induced senescence through upregulation of p16 expression has been primarily seen in RAS-mutated cells or its downstream effectors, RAF and MEK [18,19,20,21]. This senescence leads to growth arrest and a low proliferation rate in BRAF-mutated cells, an outcome that would seem to be contrary to achieving the all-important pro-proliferative and pro-survival goal of the BRAF mutation. Another function of p16 and other cyclin-dependent kinase inhibitors, e.g., p19 and p21, is to induce autophagy that is in an active and reciprocal relation with the induction of senescence [22]. Autophagy is a fundamental cellular process that eliminates molecules and sub-cellular elements via lysosomes to promote various activities, such as homeostasis, development, and differentiation [23]. Senescent cells are also characterized by several secreted factors, such as insulin-like growth factor binding protein-7 and senescence-associated secretory phenotype inflammasomes that have an impact upon tumor microenvironment (TME) cells [19, 20, 24]. Concurrently, fibroblasts undergo senescence (expression of p16) or apoptosis (expression of caspase 3), and some of them undergo transition into myofibroblasts. These senescence-related secreting factors also have an effect upon the type of inflammatory response, which is usually the immune suppressive type [25]. The main effects of BRAFi on the mutated cells consist of further inhibition of cell proliferation and survival, reversing the cytokine/chemokine profile from immune-suppressor to pro-inflammatory, and inducing cell death (Fig. 1B).

Fig. 1
figure 1

(A) BRAF mutation – main effector pathways in mutated cell and in stromal immune cells and fibroblasts; (B) BRAF inhibitor – main effector pathways in mutated cell and in stromal immune cells and fibroblasts. Abbreviations: BRAFm – BRAF mutation; CD – clone of differentiation; IL – interleukin; MCH – major histocompatibility complex; MDSC – myeloid–derived suppressor cell; MF – myofibroblast; PDGF – platelet–derived growth factor; sSASP – secreted senescence–associated secretory phenotype; sIGFBP7 – secreted insulin–like growth factor binding protein 7; TIL – tumor–infiltrating lymphocytes; Treg – regulatory T cells; VEGF – vascular endothelial growth factor

Full size image

Initial microscopic examination of the post-treated residual ameloblastomas revealed that tumors had undergone remarkable architectural and cellular changes [15]. Interestingly, most of the reports on the results following the use of BRAFi (with/without MEKi) in the management of ameloblastomas have provided only minimal information on the post-surgery microscopic findings [7, 8, 10,11,12,13, 17]. An exception is that of Tan et al., [9] who were more specific and provided details on macroscopic cyst formation, microscopic degenerative changes, squamous differentiation, and neutrophil infiltration in the ameloblastoma lining epithelium at the luminal surface of the cystic structures.

The aim of the present study was to provide what we believe to be the first comprehensive description of changes in tumor morphology and selective molecular markers in either BRAFi- or combined BRAFi-MEKi-treated ameloblastomas. The importance of this study was discussed in light of the currently available knowledge of an evolving approach for treating ameloblastomas via targeted therapy and the need for a guiding tool for clinicians in outcome assessment associated with this approach.

Methods

Between 2017 and 2023, the Department of Cranio-Maxillofacial Surgery, Chaim Sheba Medical Center, Tel Hashomer, Israel managed 14 patients diagnosed with BRAF-mutated ameloblastomas, whom completed targeted treatment and have undergone conservative definitive surgery that yielded sufficient tumor tissue for all the performed stains. Patients with large ameloblastomas that led to facial deformation were offered first line neoadjuvant therapy with BRAFi (with/without MEKi), as a jaw preservation approach (detailed protocol in supplementary information #1), instead of the standard of care surgery which would have comprised jaw resection with immediate reconstruction surgery via free vascularized tissue transfer. BRAF mutation status was performed using routine panels for standard next generation sequencing assays (NGS, Colon and Lung Cancer panel, AmpliSeq™, Thermo Fisher Scientific, Wilmington, DE, USA). Procedures included DNA extraction from paraffin sections containing areas with at least 90% tumor cells using Qiagen columns (Thermo Fisher Scientific, Wilmington, DE, USA). Ten nanograms of DNA (measured with Qubit®) were used for NGS. An amplicon library was generated for sequencing 1840 hotspot mutations in 22 genes including, BRAF, CTNNB1, KRAS, MAP2K1, NRAS, and TP53, using the “Colon and Lung Cancer panel” (Ampliseq™, Life Technologies). Raw data were analyzed using ion reporter software (Life Technologies). Each mutation was verified using Integrative Genome Viewer (IGV, Broad Institute). The BRAF mutation status alone was identified by the Biocartis Idylla system (Thermo Fisher Scientific) in three cases.

BRAFi/MEKi was provided through an expanded access compassionate program by Novartis Pharmaceuticals (Basel, Switzerland). The study was approved by the local IRB, SMC-9405-22 and performed in accordance with the Declaration of Helsinki.

The following data for all of the patients were retrieved from the institutional database of medical records: age, sex, diagnosis (conventional ameloblastomas, mural type unicystic ameloblastomas), frequency of BRAF mutation, treatment (BRAFi +/- MEKi), and treatment duration referring to time to maximal radiologic response with the greatest reduction in tumor size (nadir status), or earlier termination of the treatment due to either a lack of compliance, minimal radiological response, or adverse side effects [16]. Our definition for the microscopic response had been described in detail in our earlier report [16], and it was applied in the present study. In brief, a near-complete response was one that still displayed a few residual small, scattered clusters of ameloblastoma, while presence of changes in only part of the residual tumor alongside vital and morphologically unchanged tumor, was considered a partial response. We used a total of 7 parameters to evaluate the tumor response (especially for tumors with partial response), three of which were epithelial-related and four stroma-related, as detailed below.

Histomorphologic Features and Estimation of Changes

The histopathologic features of conventional ameloblastomas and those of unicystic ameloblastomas are well recognized [1, 2, 26] and served as a reference for analyzing BRAFi/MEKi post-treatment changes. We examined changes of the epithelial component in terms architecture (classical peripheral palisades and central loose, stellate reticulum-like layering), its cellular modifications (appearance of eosinophilic cytoplasm, pyknotic nuclei, and intercellular loss of cohesion). In addition, we assessed the presence of squamous metaplasia/differentiation with keratin formation. The extent of these epithelial changes were graded as 0 = well-conserved tumor, 1 = focal changes affecting less than 50% of the tumor, and 3 = changes affecting 50% and more of the tumor. The rationale of a 0, 1 and 3 score (rather than 0, 1, 2) was aimed to highlight the features with the most remarkable changes and discriminate between low- and high-responder tumors.

The tumor stroma was assessed for changes of fibrosis, diffuse and granulomatous type of inflammatory reaction, new bone formation/woven bone trabeculae (within tumor stroma and not at its periphery). These stromal changes were scored as 0 = none, 1 = focal/minimal, and 3 = frequent.

We used alcian blue histochemical stain (Ventana BenchMark; Roche Tissue Diagnostics, Tucson, AZ, USA) in a qualitative manner to highlight the relatively loose/myxomatous areas. Masson tri-chrome (Ventana BenchMark) was used for those tumors with post-treatment extensive fibrosis.

Whenever feasible, we compared the features of the tumors before and after treatment using the same set of parameters and scores.

Immunohistochemical Markers – Qualitative Estimation of Their Expression

Immunostaining with cytokeratin 19 was used qualitatively to emphasize the amount and extent of residual ameloblastoma in several cases. Immunostaining with p16 was also performed as it served as a marker for oncogenic activation of the mitogen-activated protein kinase (MAPK) pathway-induced cellular senescence [18, 27], and as its expression (nuclear and cytoplasmic) had been found to be associated with squamous differentiation in ameloblastomas [28]. We assessed apoptosis by means of caspase 3, using an antibody which identified its activated form as nuclear expression [29]. We highlighted macrophage infiltration in either a diffuse form or in a granulomatous pattern by immunostaining with CD68. Table 1 summarizes the technical details of the antibodies that were used.

Table 1 Antibodies used for the immunohistochemical study. All stains were performed using Ventana BenchMark Ultra immunostainer and protocols (Roche Tissue Diagnostics, Tucson, AZ, USA). Brown diaminobenzidine (DAB) staining was used to visualize positively stained cells
Full size table

Statistical Analysis

The results were presented as the mean scores of the parameters for epithelium, stroma, and total tumor (i.e., the mean of epithelium + stroma). Differences in mean scores of the epithelium and stroma between tumors with partial response and tumors with a near-complete response were analyzed by Mann-Whitney test, and between before and after treatment were analyzed by paired T-test. Spearman’s correlation test was used to analyze correlations between the duration of treatment and the various microscopic mean scores (total, epithelium, stroma, and each of the individual histomorphologic parameters). Significance was set at p < 0.05. All analyses were performed using SPSS, version 28 (IBM, Chicago, IL, USA).

Results

Study Group

The study comprised 14 patients, 11 of whom were diagnosed as having mandibular tumors (near complete response, n = 3; partial response, n = 8) and three as having maxillary tumors (all partial response). Thirteen had the BRAFV600E mutation and one with a maxillary tumor had a BRAF-SND fusion. Table 2 summarizes the patients’ clinical data, diagnosis (conventional, unicystic ameloblastoma - mural type), variation allele frequency of BRAFV600E, treatment modality (BRAFi/BRAFi + MEKi/MEKi), and duration of treatment. Variations in allele frequency of BRAFV600E before the targeted treatment ranged between 2 and 57.87%, and none were related to the response to treatment. Statistical analysis was performed for the entire study group (n = 14), for the mandibular tumors (n = 11), for the mandibular tumors with partial response (n = 8), and for the mandibular tumors with near complete response (n = 3). Six cases were also compared for their pre- and post-treatment status.

Table 2 Clinical data, treatment regimen and microscopical scores per tumor, stroma and total (tumor + stroma)
Full size table

Histomorphologic and Molecular Changes

Considering the neoplastic epithelium, architectural and cellular histomorphologic changes were the most remarkable findings in post-treatment ameloblastomas, in both conventional and unicystic, and their extent appeared to increase with the duration of treatment (Figs. 2, 3 and 4). In tumors with partial response, in addition to areas of vital, regular ameloblastoma, there were parts of tumors that lost the characteristic peripheral palisading appearance. Furthermore, neoplastic epithelial cells of the solid or cystic tumors showed various combinations of eosinophilic cytoplasm with/without granular appearance, hyperchromatic nuclei. and inter-cellular dis-cohesion. The stellate reticulum-like areas within the central areas of solid epithelial structures occasionally showed full squamous differentiation, including formation of desmosomes and keratin pearls. The greatest effect of the treatment was noted when the epithelial tumor islands were severely deformed and distorted, with only some ruptured cellular and nuclear debris as the sole evidence for the previous existence of a tumor.

Fig. 2
figure 2

Photomicrographs of treated unicystic ameloblastoma, mural type, with a near complete response. (a) collapsed cystic structure, small residual lining fragment of ameloblastoma (rectangle); (b) residual ameloblastoma from “a” at higher magnification: non-specific features, vague basal palisades; (c) p16 immunostain: positive in residual ameloblastoma (rectangle) and stroma; (d) residual ameloblastoma from “c” at higher magnification; (e) caspase 3: positive in nuclei of stromal cells, residual ameloblastoma positive cytoplasm (rectangle); (f) caspase 3, rectangle from “e” at higher magnification: no nuclear staining in ameloblastoma; (g) CD68 immunostain: positive macrophages in granulomas associated with residual ameloblastoma (rectangle); (h) CD68 immunostain: positive scattered macrophages nearby focus of residual ameloblastoma; (i) alcian blue: myxoid stroma, weak intensity also nearby residual ameloblastoma (rectangle); (j) alcian blue: area in rectangle from “i” at a higher magnification

Full size image
Fig. 3
figure 3

Photomicrographs of treated conventional ameloblastoma with partial response. (a) vital ameloblastoma; (b) ameloblastoma islands with acanthomatous/squamous changes in the center areas (asterisks); (c) right - vital ameloblastoma, left - highly dense collagenous/fibrotic stroma with non-vital ameloblastoma; (d) Masson tri-chrome stain, left: dense collagenous/fibrotic stroma, non-vital ameloblastoma; (e) collagenous/fibrotic area at higher magnification: islands of vital (asterisk) and non-vital ameloblastoma (arrows); (f) collagenous/fibrotic area at higher magnification: slit-like spaces containing tumor debris or entirely devoid of tumor (arrows), outlined by a bluish band of myxoid stroma; (g) - non-vital ameloblastoma; (h) - small new bone trabecula in fibrotic area (arrow); (i) – p16 immunostain - strongly positive tumor; (j) – p16 immunostain at higher magnification: keratin pearl formation (arrow); (k) - Caspase 3 immunostain: positive only in stromal cells (arrows); (l) - SATB2 immunostain: positive in association with a new bone small trabecula (arrow); (m). alcian blue stain: strongly stained band surrounding vital ameloblastoma; (n). alcian blue stain: weak intensity in area of new bone formation (asterisk)

Full size image
Fig. 4
figure 4

Photomicrographs of 3-week treated conventional ameloblastoma with partial response. (a) follicular area: partial loss of palisades, central areas with cystic changes, intercellular discohesion and squamous metaplasia; (b) plexiform area: columnar cells with extensive granular change indicating senescence (28); (c) follicular area: no palisading, squamous and spindle cell changes; (d) nests of distorted residual tumor (arrows); (e) p16 immunostain: highlights squamous change, including keratin pearl formation; (f) p16 immunostain, higher magnification from “e”: squamous change and keratin pearls; also positive stromal cells; (g) p16 immunostain: strongly positive in plexiform areas, abundant positive stromal cells; (h) p16 immunostain, higher magnification from “g”: strongly and uniformly positive in plexiform areas, abundant positive stromal cells; (i) CD68 immunostain: positive macrophages; (j) caspase 3: positive stromal cells; (k). alcian blue: strongly stained ameloblastoma-associated stroma, remaining stroma weakly stained; (l). alcian blue: strongly stained ameloblastoma-associated stroma in plexiform area, remaining stroma weakly/not stained

Full size image

Considering stromal changes, the inflammatory reaction was the most noticeable finding in post-treated tumors. Varying degrees of scattered chronic inflammation with or without granulomas were also apparent. Granulomas seemed to be common in areas where the tumor was no longer identified as having being replaced by a fibrotic connective tissue, which was highlighted by Masson tri-chrome stain. The lumina of cystic tumors were filled with tumor debris usually accompanied by mostly chronic inflammatory cells, either scattered or as a granulomatous type of inflammation. Myxoid niches of stroma closely surrounded the tumoral epithelium or occupied more extensive areas – these were highlighted by alcian blue stain. We could identify small trabeculae of osteoid or woven bone surrounded by plump osteoblasts within the tumor stroma in several tumors. More substantial bone formation was occasionally observed at the periphery of the tumors. The total scores for each examined parameter, epithelium, and stroma together with their mean scores are shown in Table 2; Fig. 5.

Fig. 5
figure 5

Mean scores of histomorphologic parameters in post-treated ameloblastomas. Abbreviations: Mand_NearCompl – mandibular tumors with near-complete response (n = 3); Mand_All – all treated mandibular tumors (n = 11); Mand + Max – all treated tumors, mandibular and maxillary (n = 14); Architect_changes – architectural changes; Squamous_diff – squamous differentiation; Diffuse_inflamm – diffuse inflammation; Granuloma_inflamm – granulomatous inflammation; New_bone – new bone formation

Full size image

Immunostaining revealed that the distribution of post-treatment expression of p16 varied in both the neoplastic epithelium as well as in the stroma, ranging between strong to no reaction, even within the same tumor. The expression of p16 frequently highlighted the areas with squamous differentiation but was not limited to these areas. p16 also highlighted cells with a spindle-shaped morphology within the stroma. Expression of caspase 3 in the neoplastic epithelium was almost absent, in contrast to many of the stromal cells being positive.

Duration of Treatment and Epithelium and Stroma Scores

Table 3 summarizes the mean epithelium and stroma scores and the combined (epithelium + stroma) scores for all 14 examined samples. Specifically, the mandibular tumors were subdivided further into those with partial (n = 8) and near-complete response (n = 3). The mean epithelial score of all 14 tumors in combination (2.11 ± 0.57) was higher than that of the mean stroma score (1.25 ± 0.7) (p = 0.003). This pattern was similar to the scores for all mandibular tumors (epithelium = 2.12 ± 0.51 and stroma = 1.433 ± 0.66; p = 0.035). Tumors with partial response did not differ significantly from those with near-complete response with regard to mean total epithelial + stroma, epithelial, and stroma scores (p = 0.65, p = 0.55, and p = 0.27, respectively).

Table 3 Mean scores of the epithelial parameters, stromal parameters and total (epithelium and stroma) parameters for all 14 examined tumors, for the 11 mandibular tumors, for mandibular tumors with partial response (n = 8) and tumors with near complete response (n = 3)
Full size table

The mean epithelial score of the total cohort (n = 14) did not correlate with treatment duration (r = 0.3, p = 0.28), whereas the mean total and stroma scores correlated positively with treatment duration (r = 0.64, p = 0.013 and r = 0.66, p = 0.01, respectively). Architectural changes, fibrosis, and granulomatous type of inflammation also correlated significantly with treatment duration (r = 0.533, p = 0.05; r = 0.562, p = 0.037; r = 0.672, p = 0.008, respectively).

Treatment duration of mandibular tumors (n = 11) correlated with the stroma score (r = 0.66, p = 0.025) and the total (r = 0.64, p = 0.03) but not with the epithelial score (r = 0.47, p = 0.14). Among all of the histomorphologic parameters examined in this study, the formation of a granulomatous type of inflammation emerged as being the only one significantly associated with the duration of treatment (r = 0.7602, p = 0.007).

For each patient, the total score of the epithelium was always higher than that of the total score of the stroma, irrespective of treatment duration. However as treatment was longer, the total score that the stroma achieved was higher and closer to that of the epithelium (Supplemental Information #2).

Epithelium, Stroma and Total Tumor Scores, Before Versus After Treatment

Table 4 summarizes the mean epithelial and stromal scores of all examined parameters in the six cases, for which tissue was available for comparisons. Significant differences were found regarding the epithelial architectural and cellular changes and stromal fibrosis and granulomatous type of inflammation. The overall after treatment epithelial and stromal scores were significantly higher than those before treatment (2.27 ± 0.4 and 0.64 ± 0.063, p = 0.005, and 1.66 ± 0.34 and 0.75 + 0.28, respectively, p = 0.003).

Table 4 Mean scores of epithelial and stromal parameters, separated and combined, before versus after treatment
Full size table

Discussion

The present study aimed to provide a novel description and measurable assessment of the histomorphologic modifications and selective molecular changes in BRAF-mutated, target-treated ameloblastomas. We separately scored the neoplastic epithelium and surrounding stroma and correlated these changes with treatment duration. Although preliminary, the results revealed tumor-specific changes of the neoplastic epithelium and its stroma that have occurred in what seems to be a predictable time-related manner and were not necessarily related to the type of treatment (single/combination, dose). Specifically, the neoplastic epithelium appeared to have responded to treatment first and rapidly, after which the stromal response (inflammation and fibrosis) evolved and grew larger during treatment, thus enhancing the therapeutic effect of the inhibitors – a response process that usually has developed over a prolonged period of time (weeks-to-months). Our results may raise the possibility that treated ameloblastomas carry the potential for (near) complete microscopic response but that its achievement depends upon complex and individual tumor- and patient/treatment-related factors, such as age, comorbidities, bone metabolism rate (young versus elder patients), compliance with treatment, and adverse side effects. The suggested scoring system may add insights especially for tumors with partial response in terms of assessing which compartment of the tumor has not responded as expected and aid the clinical team for planning additional adjuvant treatment options aimed at jaw preservation.

BRAFi-related histomorphologic changes have thus far been reported in malignant tumors, and these alterations seem to be tumor type-related. For example, post-treatment histomorphologic changes in melanomas are comprised of small tumor cells with eosinophilic cytoplasm and hyperchromatic and pleomorphic nuclei [18], suggesting a process of apoptosis molecularly supported by caspase 3 (a marker for apoptosis) in only a negligible fraction of tumor cells [30]. Post-treatment extensive tumor necrosis was reported on other BRAFi-treated tumors, such as anaplastic thyroid carcinoma [31, 32], and as usually being anticipated in tumors with high proliferation rates [33]. While malignant tumors carry several critical mutations that lead to high proliferation when in combination [15], ameloblastomas are heavily reliant upon the driver BRAF mutation [7,8,9, 34] that is associated with low proliferation rates. The current series comprised patients that had been treated for different periods of time, therefore providing us with the opportunity to examine microscopic changes in a time-wise manner. We found significant changes in epithelial architectural and cellular epithelial in post-treatment tumors compared to the counterpart pre-treatment status. These post-treatment findings showed unique and consistent tumor-related changes, which displayed a similar trend of changes for all treatment durations. These changes could not be attributed to apoptosis, since caspase 3 was almost entirely absent. The absence of necrosis can be explained by a low proliferation rate conferred by BRAF mutation-induced senescence, as reflected by strong and diffuse expression of p16 [35,36,37,38,39]. In terms of treatment, this low proliferation rate is a feasible explanation for the expected slow clinical response of ameloblastomas to BRAFi [8, 9, 17, 34] extending over several weeks and months, unlike malignant tumors that usually show a rapid response. In addition, treatment regimen (single/combination, dose) is not expected to change the fairly predictable time-related pattern of response in ameloblastoma.

Squamous changes were a common feature in post-treatment ameloblastomas. This finding has been similarly reported in the literature [9, 17], with the suggestion that treatment with BRAFi could have directed the tumor towards a more mature phenotype [9]. Interestingly, we observed an overlap between p16 expression and areas showing squamous differentiation, as had also been observed by others [28]. Although BRAFi agents themselves have the potential to induce senescence (upregulated p16 expression) in targeted ameloblastoma cells [40, 41], their contribution was not found to be significant when pre- and post-treatment tumors were compared. However, the joint senescence impact of the BRAF mutation and BRAFi in the setting of ameloblastoma is probably followed by an effect of autophagy, which is likely associated with further squamous differentiation of cells. Autophagy has been recently shown to occur in ameloblastomas [42], raising the possibility of a sequence of events comprised of BRAF mutation/BRAFi leading to senescence (p16 expression), autophagy, and squamous differentiation. An impact on the adjacent TME cells through BRAFi-induced senescence in the neoplastic epithelium, causing TME cells to undergo senescence (expression of p16) and apoptosis (caspase 3) [19, 20, 24], occurred in the treated tumors. We can suggest that the additional impact of BRAFi to that of BRAF mutations in this aspect was in regard to senescent fibroblasts that underwent transition into myofibroblasts, ultimately leading to significant fibrosis in the treated tumors.

The extracellular matrix of ameloblastomas had been revealed to be determined by the neoplastic epithelium that expresses factors that positively control osteoclastogenesis, such as a receptor activator for nuclear factor κappa B ligand (RANKL) and interleukin (IL)-6, on the one hand, while also expressing factors, such as secreted-frizzled-related protein-2 that suppresses osteoblasts, on the other hand [43]. This creates a TME that blocks new bone formation and encourages bone absorption. That study also found that these effects were intensified in ameloblastomas with myxoid stroma. It has been recently reported that a myxoid type of matrix may represent a state of senescence with a distinct inflammatory signature (i.e., rich in macrophages and cytotoxic T cells) as well as a rich vasculature [44]. Senescence damages bone remodeling by impairing bone formation and osteoblast progenitor cell function, thus promoting osteoclastogenesis [45, 46]. We found extensive myxoid matrix in many of BRAFi-treated ameloblastomas, as well as abundant stromal expression of p16. Collectively, we can assume that a myxoid stroma rich in senescent cells might maintain a pro-osteoclastic environment in ameloblastomas and thus could preclude bone regeneration. By eliminating the neoplastic epithelium, BRAFi reduces one pro-osteoclast-promoting factor, however, these are preliminary findings that emphasize the complexity of interactions and necessitate further research on the nature of the extracellular matrix (myxoid versus fibrotic) in the response to targeted therapy.

One of the most important known effects of BRAFi is its impact on the inflammatory response by creating an anti-tumor environment by recruiting tumor-infiltrating lymphocytes and cytotoxic lymphocytes and down-regulating immune suppressor factors (Fig. 1B) [25]. As a result, tumor cells undergo pyroptosis, which is an inflammation-associated cell death [25], rather than apoptosis. Our results are in line with this: we found a marked inflammatory reaction that targeted and destroyed the neoplastic epithelium. Lack of expression of caspase 3 indicated the absence of apoptosis, thus the final word on cell death pathways in BRAFi-treated ameloblastomas awaits further research. A macrophage-rich infiltration including a granulomatous type of inflammatory reaction, which we found to be related to the duration of treatment, seemed to represent the final step in the tumor destruction and debris clearance following BRAFi [18, 47]. The granulomatous inflammation constituted a significant change in post-treated tumors and it is in conform with documented immune-related pathologic response criteria in other target-treated tumors [48, 49].

When Pathology Meets the Clinics and Vice Versa

Collectively, the above-described morphological and molecular changes in BRAFi-treated ameloblastoma were specific for this tumor and likely dependent upon built-in features. We can now assume that these tumors, especially those in the mandible, followed a predictable pattern of behavior with emphasis upon a relatively slow reaction to treatment, irrespective of treatment regimen, a feature which should be taken into consideration by the clinicians. Other potential factors that could affect the time and extent of response are patient-related, as already mentioned, which should be discussed by a multidisciplinary team.

The case for a conservative approach for ameloblastomas versus resection with safety margins of 1.5 cm beyond the radiological margins has already and eloquently been debated throughout the pre-BRAFi era [50]. The stated reasons included the disproportionate number of these tumors occurring in young patients, the subsequent impact on quality of life subsequent to a radical operation (all patients), and a 15–25% rate of recurrence [51, 52]. According to our findings, cases with a near-complete pathological response retained minimal tumor residue that could be successfully removed conservatively, highlighting the need for a meticulous follow-up protocol. In regard to those tumors with partial response, analyzing microscopic findings according to the series of suggested parameters, should aid the treating team in understanding which tumor component has responded or failed to respond, and serve as a basis for decision taking toward subsequent steps in adjuvant treatment planning (e.g., follow up, conservative surgery, modifications/changes in targeted treatment regimen, combinations of these approaches). This is especially pertinent to mandibular ameloblastomas, even those with a large recurrent lesion, for which BRAFi treatment could be reinstated with a substantial likelihood of success [13], since these tumors are driven by one mutation with almost no other allied mutations and a low risk for the tumor to escape from targeted therapy.

Study Limitations

One limitation of this study is the small number of cases for comparing the morphological and molecular features between the original and post-treated tumors in all patients. Since some of the patients were referred to us from other national and international medical centers, the paraffin blocks of the original tumors were not always available for molecular evaluation. Histomorphologic parameters and molecular markers examined in this initial study were selected on the basis of their being connected to known effector pathways of the BRAF mutation and impacts elicited by BRAFi, without previous knowledge of what to expect. It is to be assumed that comparing a higher number of pre- and post-treatment specimens will pave the way to update the presented panel of parameters and focus on those with immediate clinical application. Collaborative international studies with larger numbers of targeted-treated ameloblastomas are likely to enable the establishment of clinical, histopathologic, and molecular databases, which will ultimately aid in creating new guidelines for the treatment of this tumor.

Conclusion

Treatment of ameloblastoma with targeted therapy is predictable and time related, enabling jaw preservation treatment strategy. Unlike malignant soft tissue tumors, when treating intra bony ameloblastomas, several months of treatment duration is expected due to the tumor’s slow clinical response and the associated bone metabolism. Time-wise, the reaction of the neoplastic epithelium is first and rapid, while the stromal response (inflammation and fibrosis) evolves and increases during treatment, the entire process expanding over weeks-to-months. Post surgical pathological evalutaion of the tumor tissue is essential as it can highlight which tumor components have responded or failed to respond. This should serve as the basis for the clinical decision on the need for adjuvant treatment, with a prime aim of jaw preservation, prevention of reccurent disease and minimizing sequelae.