process by which a cavity comes to be lined by the proliferating odontogenic epithelium. Many will think it astonishing that the actual process of cyst formation is so poorly understood, and although a number of theories have been debated for decades, there are very few observational or experimental data to support or refute any of them. There are three proposed mechanisms of cyst formation (Shear 1963a ; Summers 1974 ; Valderhaug 1974 ; Lin et al. 2007 ; Nair et al. 2008 ; Huang 2010 ):
1 Central necrosis or nutritional deficiency theory. This proposes that a cyst cavity forms within a proliferating mass of epithelial cells due to loss of nutrition followed by degeneration and death of cells in the centre of the mass.
2 Abscess theory. This postulates that the proliferating epithelium surrounds an abscess cavity, in effect walling off the central focus of inflammation. Essentially this represents the process of normal wound healing, where epithelium proliferates to cover denuded connective tissues.
3 Merging epithelial strands theory. This proposes that the proliferating epithelium forms a three‐dimensional ‘ball mass’ (Lin et al. 2007 ), which entraps inflamed connective tissue. This connective tissue then breaks down due to loss of a blood supply and a cyst cavity forms.
Proponents of each process often promote their favoured theory, while suggesting that others are not tenable (Lin et al. 2007 ; Nair et al. 2008 ; Huang 2010 ), but in fact they are not mutually exclusive and all three have a similar premise – that epithelium proliferates to surround and encase a focus of inflamed and degenerating or necrotic tissue. There is little difference, for example, between an abscess (theory 2) and necrotic connective tissue (theory 3), apart perhaps from the degree or stage of necrosis and the types of cells present. Similarly, the necrotic centre of an epithelial mass (theory 1) may contain PMNs and essentially become an abscess, a feature familiar to tumour pathologists in the form of comedo necrosis. Thus on histological examination, the presence of an abscess cavity lined by epithelium may support theory 1 or 2. Also, in the previous edition of this book we showed a photomicrograph (Figure 3.7) of arcading epithelium surrounding inflamed connective tissue in support of theory 1 – but in fact this pattern of proliferating epithelium, with entrapment of connective tissue, is very similar to the merging epithelial strands theory of Lin et al. (2007 ) (theory 3).
Figure 3.7 Arcades and rings of proliferating epithelium in a periapical granuloma. The epithelium surrounds islands of inflamed connective tissue that may break down to initiate early cyst formation.
In support of theory 1, there is some histological evidence for central necrosis or nutritional deficiency. In some periapical lesions, distinct clefts may be seen in sheets of epithelial cells (Shear 1963a ) (Figure 3.8), and the proliferating epithelial masses often show considerable intercellular oedema. These intercellular accumulations of fluid coalesce to form microcysts containing epithelial and inflammatory cells, including PMNs (Figure 3.9). These central areas have been shown to contain high levels of acid phosphatase activity and proteolytic enzymes consistent with autolysis or necrosis (Lutz et al. 1965 ; Summers 1972 , 1974 ). Microcysts may increase in size by coalescence with adjacent microcysts and, once established, the cyst increases in size by mechanisms that are discussed later.
Most information regarding cyst formation is taken from histological observation of biopsy specimens taken from humans and there are very few longitudinal studies. In experiments that have not been repeated, Valderhaug (1972 ) induced radicular cysts in monkeys. Because of the rarity of such experiments, it is worth considering some of the details of his findings. He induced pulpal necrosis sequentially in 39 teeth in 4 animals and was able to histologically examine periapical inflammation and cyst formation for up to 360 days after the pulps were removed. Of interest is that he did not observe any cysts until after 200 days, although proliferating epithelium was observed in earlier lesions. In lesions examined after 200 days, 11 of 16 (69%) developed cysts. Even after 300 days, 3 of 8 lesions showed no inflammation, but 4 had developed into cysts. Of relevance to the proposed mechanisms of cyst formation is that he did not see evidence of intraepithelial degeneration with formation of microcysts, but observed long strands of proliferating epithelium lining surfaces of granulation tissue, or arcades surrounding cores of vascularised or degenerating granulation tissue. These observations support the merging epithelial strands theory (theory 3). However, in most cases Valderhaug also observed that the proliferating epithelium was associated with PMNs, but he did not describe frank abscess formation. He also showed that in most cysts the epithelial lining was closely connected to the roots around the apical foramen, supporting the notion that the epithelium is reforming an intact integument, and that the cysts are pocket or bay cysts. In a very similar study, Valderhaug (1974 ) examined 52 primary teeth, and although periapical inflammation was common, he found small cysts only in ‘a few cases with long observation periods’. Of relevance to the theories of cyst formation is that he found that abscess formation, often with oral fistulas, was common and more frequently seen than in his previous study on permanent teeth (Valderhaug 1972 ). His findings, however, did not support the abscess theory, since ‘proliferating epithelium was not observed in the periapical area in connection with abscess formation’.
Figure 3.8 Sheet of epithelial cells in a periapical lesion. A distinct cleft has formed and this may initiate a radicular cyst.
Despite this, as described previously, there is good evidence that PMNs are a prominent feature of periapical granulomas and are associated with epithelial proliferation (Valderhaug 1972 ; Shear 1963a , 1964 ; Cohen 1979 ; Johannessen 1986 ; Marton and Kiss 2014 ). In their examination of 256 periapical lesions, Nair et al. (1996 ) showed that 90 (35%) were an ‘abscess’ and that two‐thirds of these contained epithelium. However, the definition of an abscess was of a collection of PMNs in a pre‐existing periapical granuloma – making distinction of a primary abscess from a collection of PMNs in a necrotic lesion difficult. In support of the abscess theory, Nair et al. (2008 ) were able to demonstrate that epithelium implanted into experimentally induced abscesses could form cysts. However, the relevance of these experiments is uncertain, since the abscesses were induced in the skin of experimental rats, and cysts only developed in 2 of 16 animals (6%). In addition, it is known that implanted epithelium may cause cysts even in the absence of infection or abscess formation. Nevertheless, we have seen cases of abscess formation in periapical granulomas where the abscess cavity has become encased in epithelium to form a cyst filled with PMNs (Figure 3.10). This supports the abscess theory, but not to the exclusion of the other possible mechanisms.
Figure 3.9 Degeneration of cells in the centre of a mass of proliferating epithelium in a periapical granuloma. There is an intense infiltration of lymphocytes and polymorphonuclear leukocytes. Accumulations of intercellular fluid coalesce to form a microcyst.
It seems, therefore, that there is little difference between the three proposed mechanisms – cyst formation occurs due to a ‘walling‐off’ of inflamed connective tissue by a process of epithelial proliferation similar to healing at an epithelial surface. This is due to the innate property of epithelium to form an external protective integument. The cyst lumen therefore represents the external environment, and in the case of a pocket