More recently, Maruyama et al. (2015 ) found that inflammatory collateral cysts consistently expressed cytokeratins K13, K14, and K19, and perlecan and UEA‐1. Dentigerous cysts showed similar expression, but also expressed K10, which was negative in the collateral cysts. Other cyst types (odontogenic keratocyst and radicular cyst) showed minor differences in keratin expression, but the expression of the markers in lateral periodontal cyst and in normal junctional or sulcular epithelium was similar to the collateral cysts. Overall, the expression of cytokeratins was variable and the results of this study cannot be regarded as conclusive. Nevertheless, the similarity of expression patterns between the inflammatory collateral cysts and junctional/sulcular epithelium supports the view that the cysts arise from the epithelium lining a periodontal or pericoronal pocket. The authors suggest that their results contradict the findings of Craig (1976 ), who they say proposed an origin from ‘reduced enamel epithelium’, but the findings actually support Craig's proposal that the epithelium of origin is from the early epithelial attachment (i.e. the junctional or sulcular epithelium) that derives from the reduced enamel epithelium.
An origin from pocket or pericoronal epithelium is further supported by the observation that the cyst lining is often continuous with junctional or sulcular epithelium, or with the periodontal or pericoronal pocket around the associated tooth (Figures 4.6 and 4.7; Craig 1976 ; Ackermann et al. 1987 ; de Sousa et al. 2001 ; Colgan et al. 2002 ; Philipsen et al. 2004 ). The cyst therefore presents as a dilated pocket, whereby a probe can be placed into the lumen through the opening of the associated periodontal pocket or from beneath the operculum (Figure 4.6b). Some workers require an opening to the surface as a diagnostic criterion (de Sousa et al. 2001 ). A number of workers (Craig 1976 ; Ackermann et al. 1987 ; de Sousa et al. 2001 ) have obtained sections of the cyst in continuity with the associated teeth and showed that the cyst linings were attached at the cementoenamel junction and were continuous with the oral epithelium (Figure 4.6).
Figure 4.6 Gross specimen of a paradental cyst on the buccal aspect of a partially erupted third molar, which has been received intact. (a) The cyst is attached to the cementoenamel junction and involves the distal and buccal surfaces of the roots. (b) Careful probing shows that the lumen is open towards the coronal aspect and is in continuity with the lining of the pericoronal pocket.
Source: Courtesy of Prof G.T. Craig.
These studies suggest that inflammatory collateral cysts are equivalent to a dilated follicle or pocket lined by hyperplastic and proliferative epithelium derived from reduced enamel (follicular) epithelium. Thus, a descriptive designation of ‘inflammatory pocket cyst’ may be appropriate, and Slater (2003 ) has suggested that the third molar lesions should be called ‘eruption pocket cysts’. It is possible that swelling associated with inflammation leads to occlusion of the opening of the pocket, thus allowing accumulation of debris and cyst growth by osmotic pressure in a similar process to that described for radicular cysts.
From this discussion it can be seen that there is agreement that collateral cysts are of inflammatory origin and that the initiating factor is inflammation within the pericoronal tissues of an erupting or partially erupted tooth. However, pericoronitis is common and this pathogenic process does not easily explain why paradental and mandibular buccal bifurcation cysts are so rare.
Craig (1976 ) was able to examine the teeth in 28 of his 49 cysts. In 20 of the 28 cases (71.4%) he found a developmental enamel projection or spur extending from the cementoenamel junction towards the bifurcation of the roots on the buccal aspect of the tooth. He suggested that the downwards extension of the reduced enamel epithelium at the site of an enamel spur may provide a site of stagnation or focus of inflammation, and predispose to the development of a paradental cyst at this site. Fowler and Brannon (1989 ) found enamel projections on two of the three teeth they were able to examine and agreed with Craig that such projections may localise inflammation and cyst formation on the buccal aspect. In their study, Ackermann et al. (1987 ) were only able to examine eight of the associated teeth. In all cases the cyst was attached at the cementoenamel junction, and two showed enamel spurs. Others, however, have been unable to show an association between these cysts and enamel spurs, but often this is because the associated teeth were not available for examination (Vedtofte and Praetorius 1989 ; de Sousa et al. 2001 ). Studies have shown that enamel projections may be seen on up to 70% of teeth and are more common on mandibular molars (Chan et al. 2010 ). Risnes (1974 ) found that 11.7% of third molars had enamel projections and that 99% occurred on the buccal aspect. These data may explain in part why lesions are found predominantly on the buccal aspect of mandibular molars.
Colgan et al. (2002 ) suggested that food impaction may have an important part to play. In 13 of their 15 cases the associated tooth was opposed by a maxillary molar and they proposed that the angulation of the affected tooth (usually distal) could promote food impaction into the pericoronal tissues around the crown. As further evidence for this they showed that four cases contained giant cells consistent with a foreign body reaction.
With regard to the mandibular buccal bifurcation cyst, studies have shown that the affected tooth is almost always tilted buccally (Figure 4.4), giving prominence to the lingual cusps and often associated with increased pocket depth on the buccal aspect (Pompura et al. 1997 ; Philipsen et al. 2004 ). Stoneman and Worth (1983 ) suggested that the mesio‐buccal cusp of the first molar is the first to penetrate the oral mucosa during eruption, and that this would explain the buccal location of the cyst, but this explanation is at odds with the fact that the lingual cusps may erupt first in the buccally displaced tooth. However, it is not known whether this buccal inclination occurs as a result of displacement of the tooth by the cyst, or is present at eruption and may thus predispose to cyst formation. Since the cysts are rare, it is most likely that the associated tooth is buccally displaced at eruption and that this predisposes to inflammation in a buccal pocket and subsequent cyst formation. It is possible that the prominence of the lingual cusps and the buccal inclination may predispose to food impaction on the buccal aspect.
Relationship of the Paradental Cyst to the Dentigerous Cyst
As discussed previously (see ‘Frequency’), there is evidence that some clinicians and pathologists do not recognise the paradental cyst as an entity, but rather regard it as a variant of dentigerous cyst. Ackermann et al. (1987 ) suggested that although the paradental cyst may arise from follicular (reduced enamel) epithelium, the histogenesis is quite different and it should not be regarded as a variant of dentigerous cyst. They believed that the dentigerous cyst should be defined as a cyst enclosing the crown of a completely unerupted tooth.
Despite this, many clinicians and pathologists diagnose collateral lesions associated with partially erupted third molars as ‘inflamed dentigerous cysts’. Fowler and Brannon (1989 ) agreed with Ackermann et al. (1987 ) that a dentigerous cyst is, by definition, associated with an unerupted tooth, but also believed that the paradental cyst is a variant of dentigerous cyst. In this respect, a relationship to the dentigerous cyst may be suggested on the basis of a common origin from epithelium lining the dental follicle, with intraosseous (dentigerous cyst) and extraosseous (paradental cyst) variants associated with totally unerupted and partially erupted teeth, respectively.
One possible pathogenic mechanism for the paradental cyst is that it represents a developmental dentigerous cyst that has become laterally (and buccally) displaced by the eruption of the associated tooth, with subsequent inflammation. This seems unlikely, however, since the typical radiology of the paradental