Shear's Cysts of the Oral and Maxillofacial Regions. Paul M. Speight

the designation as a neoplasm was based on its ‘aggressive, infiltrative behaviour’ and the role of the PTCH gene in its pathogenesis. However, keratocysts do not meet the criteria for neoplasia suggested by Willis, since there is no evidence of uncoordinated independent growth. It is ironic that since the lesion was designated as a neoplasm in 2005, there have been almost 50 publications reporting marsupialisation (or decompression) as an effective treatment compared to only 10 papers before 2005. A number of publications have shown that marsupialisation is effective and that it may result in regression of the typical keratocyst lining to an epithelium indistinguishable from that seen in normal oral mucosa. This cannot be regarded as the behaviour of a neoplasm.

A further consideration is the use of the term ‘aggressive’. It is common for authors to assert that the odontogenic keratocyst is aggressive because it has a propensity for recurrence. While there is no doubt that the keratocyst (and a number of other cysts) has a tendency to recur, we do not support the well‐used assertion that ‘recurrence’ and ‘aggressive’ are synonymous. The recurrence rate is almost entirely associated with the pattern of cyst growth and the type of treatment (see Chapter 7) and is similar to that seen in other cysts that have a multilocular growth pattern (botryoid odontogenic cyst, glandular odontogenic cyst). Inadequate removal may lead to residual cyst elements being left in the tissues and these may proliferate to form new cysts. It is also true that occasionally a keratocyst, especially a recurrent lesion, may spill into the soft tissues and cause considerable clinical problems. However, other cysts that are known to be difficult to remove, and may have very high recurrence rates, are not described as aggressive or thought to be neoplastic. For example, inadequately treated ranulas have a recurrence rate of up to about 65% (see Chapter 15) and thyroglossal duct cysts recur in up to 50% of cases if residual ductal elements are not also removed (see Chapter 18). We believe that the term ‘aggressive’ is overused and may be over interpreted to suggest an invasive lesion or a behaviour similar to that associated with truly aggressive lesions such as malignancies. In this context, none of these cysts behaves in a way that can be described as truly aggressive.

      A more compelling argument for the neoplastic nature of cysts is the finding of genetic aberrations. A key point in this discussion, as already mentioned, is the finding of PTCH gene alterations and activation of the HH signalling pathway in a range of developmental cysts. The finding of a single point mutation or LOH in the PTCH gene is insufficient to designate the keratocyst as a neoplasm. However, more recent studies (Stojanov et al. 2020 ) have confirmed earlier work that has shown that some keratocysts show biallelic loss of both copies of the PTCH gene (Levanat et al. 1996 ; Sun et al. 2008 ; Pan et al. 2010 ). In their study, Stojanov et al. (2020 ) found that 80% of sporadic keratocysts showed biallelic loss of PTCH1. Biallelic loss of the gene means that both copies have been lost or mutated and meets the ‘two‐hit’ hypothesis of Knudson (Knudson 1971 , 1996 ), which suggests that loss of both copies of a tumour suppressor gene causes neoplasia. However, Knudson's hypothesis related to hereditary cancer (retinoblastoma), and it is now thought that two hits may not always cause neoplasia and that further genetic aberrations may be necessary (Tomlinson et al. 2001 ).

      Notwithstanding this argument, the finding of biallelic loss of the PTCH gene in some keratocysts provide some evidence that at least a subset of odontogenic keratocysts may be neoplastic. However, the line of demarcation between aberrant development and neoplasia is blurred and it seems that the keratocyst may sit on a spectrum between a developmental cyst and a benign cystic neoplasm. At the present time it is difficult to make a clinicopathological distinction between cysts showing different genetic changes and there are no defining criteria for which, if any, keratocysts might be neoplastic. Further research is needed to clearly define neoplasia and to determine if there are any behavioural differences between cysts with different genotypes or clinicopathological phenotypes.

      These issues and the debate about the neoplastic nature of the odontogenic keratocyst are discussed in detail in Chapter 7.

      Other lesions that sit on the cyst–tumour interface include the calcifying odontogenic cyst and the glandular odontogenic cyst. In 2005, the WHO renamed the calcifying odontogenic cyst the calcifying cystic odontogenic tumour, and designated all odontogenic ghost cell lesions as neoplasms (Barnes et al. 2005 ; Prætorius and Ledesma‐Montes 2005 ). However, more than 85% of ghost cell lesions are simple cysts and do not recur, with little evidence of behaviour that would suggest a neoplastic origin (Ledesma‐Montes et al. 2008 ). It has been shown, however, that ghost cell lesions, including calcifying odontogenic cysts, are characterised by mutations in the β‐catenin gene (CTNNB1) and may be driven by activation of the WNT signalling pathway (reviewed by Gomes et al. 2019 ). The presence of this CTNNB1 mutation has led to suggestions that the calcifying odontogenic cyst is a neoplasm. However, this mutation and the resulting intracellular expression of β‐catenin are seen in a range of lesions and a single mutation is not sufficient to justify designation as a neoplasm. Further research is needed, but at the present time there is little evidence to suggest that the calcifying odontogenic cyst behaves as a neoplasm.

      The relationship between the glandular odontogenic cyst and neoplasia has been speculative and is based largely on its histological similarity to intraosseous mucoepidermoid carcinoma. This is discussed in detail in Chapter 10. This histological similarity can make it very difficult to reach a definitive diagnosis and may result in misdiagnosis of lesions. A number of papers have reported lesions with histological features of a glandular odontogenic cyst that have recurred as mucoepidermoid carcinomas, but some of these are poorly illustrated and in others the diagnostic criteria are not clear. Nevertheless, this has led to suggestions that the glandular odontogenic cyst may be a precursor lesion, or a ‘benign variant’ of mucoepidermoid carcinoma. The debate is further fuelled by the observation that the glandular odontogenic cyst has a high recurrence rate, leading some authors to label it ‘biologically aggressive’ (Greer et al. 2018 ). Mucoepidermoid carcinoma is characterised by specific rearrangements of the MAML2 gene and large studies have shown that glandular odontogenic cysts do not show this change, suggesting that the two lesions are distinct and that the cyst is not a precursor to the carcinoma (Bishop et al. 2014 ). More recently, two studies have reported lesions with the features of glandular odontogenic cyst that have recurred as MAML2 positive mucoepidermoid carcinomas, and have suggested that this is further evidence that mucoepidermoid carcinomas can arise from glandular odontogenic cysts (Greer et al. 2018 ; Nagasaki et al. 2018 ). In both cases, an alternative explanation is that the primary lesions were also mucoepidermoid carcinomas, but were either misdiagnosed or were histologically indistinguishable from cysts. Taken together, however, these studies raise a number of questions and provide sufficient evidence to justify further research into the relationship between glandular odontogenic cyst and mucoepidermoid carcinoma. The two possibilities to be explored are that glandular odontogenic cyst is a benign variant or precursor of mucoepidermoid carcinoma or, more likely, that in some cases the two lesions are histologically indistinguishable. Until these issues are clarified, care must be taken to use strict criteria for diagnosis and histological assessment must be accompanied by careful assessment of the clinical and radiological features (see Boxes 10.1 and 10.2, Tables 10.3 and 10.4).

An Approach to Diagnosis of Cysts of the Jaws

An accurate diagnosis of a cyst requires knowledge, common sense, and a logical approach to the assimilation of all the information available about each lesion. With few exceptions, it is not possible to reach a correct diagnosis without considering the clinical, radiological, and histological features. This is especially relevant to histopathologists, who may be called upon to make a diagnosis on a small biopsy. To do so without considering