At this time, well over 100 cases of GOC have been reported in the English literature. Many of these have been single case reports; however, several short series and detailed reviews of the literature have also been published [1
]. Therefore, the GOC, though rare, is now relatively well known, especially among oral and head and neck pathologists. Moreover, the microscopic features of GOC have been well documented, and the most recent World Health Organization classification includes a definition of the GOC and lists numerous characteristic microscopic features of this cyst [3
]. However, it is not uncommon in the practice of pathology to encounter jaw cysts that display some but not all the features that have been described in the “classic” GOC. This situation often occurs when the cyst is associated with an unerupted tooth. It is well known that odontogenic cysts, particularly dentigerous cysts, may show metaplastic changes in the cyst lining that overlap with microscopic features described in GOC, such as eosinophilic cuboidal cells, mucous cells, and ciliated cells. [27
] Therefore, in such cases, the pathologist may be unsure whether the cyst in question is a true GOC or simply a dentigerous cyst mimicking a GOC. To our knowledge, the question of how many and which microscopic features are necessary for diagnosis of GOC has been previously addressed by only one group of investigators. Kaplan et al. [11
] proposed a list of major and minor microscopic criteria for GOC based on the frequency of each feature in reported cases from the literature. Based on their analysis, it was suggested that at least focal presence of each of the following major
criteria must be present for diagnosis:
- Squamous epithelial lining, with a flat interface with the connective tissue wall, lacking basal palisading
- Epithelium exhibiting variations in thickness along the cystic lining with or without epithelial “spheres” or “whorls” or focal luminal proliferation
- Cuboidal eosinophilic cells or “hobnail” cells
- Mucous (goblet) cells with intraepithelial mucous pools, with or without crypts lined by mucous-producing cells
- Intraepithelial glandular, microcystic, or duct-like structures
They also listed the following minor
criteria, which support the diagnosis, but are not mandatory:
- Papillary proliferation of the lining epithelium
- Ciliated cells
- Multicystic or multiluminal architecture
- Clear or vacuolated cells in the basal or spinous layers
Although these diagnostic criteria have merit, our findings demonstrate several differences with regard to specific microscopic features necessary for diagnosis of GOC. As shown in Table , two of our cases diagnosed as GOC did not contain microcysts, five did not display variable thickness of the lining, only 71.7% contained mucous cells, and only 67.4% contained epithelial spheres. Therefore, we do not believe that all of Kaplan and colleagues’ “major criteria” need be present for diagnosis. Our findings suggest it is likely a combination of specific microscopic features, not necessarily corresponding with their major and minor criteria, that appear to be important in making an accurate diagnosis of GOC. By comparing which microscopic parameters were statistically more often present in cases we diagnosed as “GOC” versus those we diagnosed as “non-GOC” (Table ), it can be seen that eosinophilic cuboidal (hobnail) cells, though necessary for diagnosis, are of little value in problematic cases, because they were also seen in 95.2% of GOC mimickers. Likewise, because mucous cells were seen in 71.7% of GOCs and 61.9% of GOC mimickers, they also are not helpful in discrimination. However, the presence of microcysts, epithelial spheres, clear cells, variable thickness of the cyst lining, and multiple compartments do appear to be of value in separating GOCs from GOC mimickers. The stratified analysis demonstrated that clear cells were marginally significant and multiple compartments were significant only when microcysts were present, while epithelial spheres were significant regardless of whether microcysts were present or not. When comparing only cases of GOCs and non-GOCs that were associated with unerupted teeth (dentigerous relationship), the presence of microcysts, clear cells, and epithelial spheres appears to be helpful in distinguishing GOCs from GOC mimickers (Table ). Table demonstrates that the presence of 7 or more microscopic parameters was highly predictive of a diagnosis of GOC in our series as only 3 of 46 cases we diagnosed as “GOC” contained less than 7 parameters. Similarly, the presence of 5 or less microscopic parameters was highly predictive of a non-GOC diagnosis in our series as only 3 of 21 cases we diagnosed as “non-GOC” contained more than 5 parameters.
As indicated in Table , two cases that displayed 8 of the 10 microscopic parameters were not diagnosed as GOC. One of these cases was a large expansile multilocular radiolucent lesion in the left anterior mandible that recurred 2½ years after initial treatment. The original cyst did not display microcysts or variable thickness of the lining. The recurrent cyst did not demonstrate microcysts, epithelial spheres, mucous cells, or variable thickness. The second case was a 4.5 cm expansile multilocular radiolucency in the anterior maxilla that by history possibly represented recurrence of a GOC removed 10 years previously (original material not available). This cyst did not contain microcysts or epithelial spheres and was significantly inflamed. While the clinical and microscopic findings raise the possibility that these cysts were actually GOCs, none of the reviewers classified them as such on initial or subsequent review. It could be argued that these two cases should not have been included in the study; however, doing so would not significantly change the statistical comparison, and we wanted to test our diagnostic criteria on all cases submitted by the study participants. It should also be emphasized that recurrent GOCs sometimes do not exhibit as many of the microscopic features as the original cyst. This may sometimes hinder the interpretation of recurrent GOCs if the original material is not available for review.
Table indicates that 8 cases of GOC were associated with an unerupted tooth (dentigerous relationship). This was an unexpected finding because the majority of reported cases of GOCs are not associated with unerupted teeth, as only 6 such cases have been previously reported [6
]. Moreover, because some of the microscopic features of GOC are similar
if not identical
to metaplastic changes in dentigerous cysts, rendering a diagnosis of GOC for a cyst in a dentigerous relationship should be done with caution. It is well-known that eosinophilic cuboidal cells, cilia, and mucous cells are occasionally seen in the lining of dentigerous cysts. Moreover, dentigerous cysts sometimes contain vacuolated areas in the cyst lining (“pseudomicrocysts”), as shown in Fig. . The cells lining these pseudomicrocysts tend to have a more flattened appearance rather than the cuboidal to columnar cells lining true microcysts. We do not believe such structures are true microcysts, and they do not satisfy the criteria for microcysts as defined previously. Within the context of these diagnostic pitfalls, we were strict in our interpretation of these 8 cases as evidenced by the fact that 3 of the cysts contained 8 of 10 parameters, 4 cysts contained 7 of 10 parameters, and 1 cyst contained 6 parameters. Also, all of the microscopic features identified in these cases were more than just focal findings. Unfortunately, no follow-up was available on 6 of the 8 cases. The remaining 2 cases showed no evidence of disease 18 months and 5½ years after initial treatment.
Vacuolated area (“pseudomicrocyst”) in lining of a dentigerous cyst (H&E stain, orig. mag × 400)
Because all of the cysts in this study were treated with conservative excision, no comparisons could be made relating type of treatment to recurrence. However, with a recurrence rate of 50%, one could argue that more aggressive treatment may be indicated for these cysts. Most cases reported in the literature have also been treated conservatively and a recurrence rate of 30% has been reported [11
]. Eighteen reported cases have been treated with more aggressive initial surgery to include en bloc and segmental resection [6
], and recurrence has been documented in only one case after such treatment [8
]. Radiographic evidence of possible recurrence was also reported in one case [31
In 3 of our cases, islands resembling central mucoepidermoid carcinoma (CMEC) were noted in the cyst wall (Fig. ). In two of these cases, the MEC-like islands invaded bone. Otherwise, the cases were “classic” GOCs microscopically. Follow-up was available on only 1 of these 3 cases, indicating no evidence of disease 6 years and 2 months after initial conservative treatment. Although it appears, based on current knowledge and experience, that these MEC-like islands in the cyst wall likely have no clinical significance, this finding suggests at least the possibility that GOC and CMEC may be related or that CMEC may arise in GOC. The possible relationship of GOC and CMEC has been previously discussed by several investigators [21
]. However, definitive evidence in this regard will have to await further study. Of interest is the discovery within the last few years that most MECs exhibit a t(11:19)(q21:p13) translocation resulting in the MECT1:MAML2 fusion protein [41
]. Recently, this translocation has also been reported in CMEC [45
]. Therefore, molecular assays specifically targeting the MEC-like islands in the cyst wall of some GOCs may aid in answering the question of whether these islands represent true malignant transformation or not.
Fig. 9 a Mucoepidermoid carcinoma-like islands in cyst wall of glandular odontogenic cyst. (Hemosiderin laden macrophages are also noted) (H&E stain, orig. mag × 200). b Mucoepidermoid carcinoma-like islands in cyst wall of glandular (more ...)