Orthokeratinized odontogenic cysts: a Spanish tertiary care centre study based on HPV DNA detection

Following a histological review of these 115 cases, 16 cases of OOC were selected based on the morphological criteria established by Wright [4]. Lesions in which all or a predominant portion of the epithelial lining showed non-corrugated orthokeratinization, with the presence of a granular layer, were included in this series. The presence of this orthokeratinization on the epithelial surface as well as of a granular layer in the thickness of the epithelium was clearly demonstrated with PAS staining in all selected cases. Likewise, we also included lesions in which basal and parabasal cells were not prominent and did not palisade or polarize. Cystic lesions containing only focal OOC areas or various skin appendages (i.e. hair follicles and sebaceous and sweat glands) were excluded.

Clinical data, including age, sex, lesion location (maxilla or mandible) (anterior, premolar or molar regions), radiologic features, surgical procedures, and information on recurrence (including optical and immunohistochemical data obtained from the recurrent OOC samples), were reviewed.

Our immunohistochemical study, performed on recurrent OOC samples included the analysis of two well-known markers (p53 and Ki67), both of which are implicated in tumorigenesis and cell proliferation [6]. Briefly, 5-μm sections were cut from the original paraffin-embedded blocks and mounted on poly-L-lysine-coated glass slides prior to immunohistochemical staining, performed using monoclonal antibodies: mouse antihuman p53 (clone DO-7, dilution 1:50, Dako®, Glostrup, Denmark) and antihuman Ki67 (clone MIB-1, dilution 1:100, Dako®, Glostrup, Denmark). Immunostaining was visualized using the high-pH EnVision FLEX system (Dako®, Glostrup, Denmark); hematoxylin was used for counterstaining and for both techniques tonsil sections with oropharyngeal epithelium were employed as positive staining controls and the negative controls were mock-stained test sections (the primary antibody was replaced with PBS).

After histological and clinical revision, formalin-fixed, paraffin embedded (FFPE) blocks of all selected OOC cases, including the one recurrent OOC, were retrieved from the archives of the Department of Pathology. The FFPE blocks were cut into three 5-μm sections, and DNA extraction was performed using a MagCore® Super HF16 automated nucleic acid extractor (RBC Bioscience Corp., New Taipei City, Taiwan).

For HPV detection and genotyping, an Anyplex II HPV28 (CE-IVD) kit (Seegene, Seoul, South Korea) was used to identify 14 high-risk and 14 low-risk HPV genotypes in two PCR reactions: mixture A (types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68) and mixture B (types 6, 11, 26, 40, 42, 43, 44, 53, 54, 61, 69, 70, 73 and 82), respectively. This semi-quantitative multiplex real-time PCR Anyplex II HPV28 system uses both dual priming oligonucleotides and tagging oligonucleotide cleavage and extension technology that combines five dyes and seven different melting-curve temperatures to distinguish the 28 HPV genotypes. PCR reactions were performed on a CFX96 real-time thermocycler (Bio-Rad Laboratories, Hercules, CA, USA) using 5 μL DNA for both mixtures A and B.

As internal control, the L1 gene of HPV DNA was simultaneously co-amplified with the human beta-globin housekeeping gene in the same PCR reaction tube. In addition, two positive controls (high-risk and low-risk HPV genotypes) were included in each PCR run along with two negative controls (non-template control) from the DNA extraction step and PCR experiment. Finally, all data from the PCR runs were interpreted using Seegene viewer software supplied by the manufacturer (Seegene, Seoul, South Korea).

The histological review of 115 keratinizing jaw cysts detected only 16 OOCs (13.91%). The cases were accepted for inclusion as OOC if they fulfilled the criterion of a complete or predominant epithelial lining with non-corrugated orthokeratinization and the presence of a granular layer. Of these 16 patients, 10 were male and 6 were female (male-to-female ratio, 1.66:1). Age at diagnosis ranged from 13 to 65 years (average 36.06 ± 13.16 years), with a clear predilection for the fourth decade of life (50%) in both sexes (Fig. 1). The mandible was affected in 12 cases (75%) and the maxilla in 4 (25%). Molars (65%) and premolars (16%) were the most commonly affected regions. Only one case showed bilateral involvement of two mandibular cystic lesions, which were diagnosed simultaneously. Of the 16 OOCs, 9 (56.5%) were associated with an impacted tooth, of which 6 were interpreted clinically as dentigerous cysts.

Jaw swelling was the most common symptom (12 cases; 75%). Four (25%) patients also complained of pain. The evolution of the symptoms ranged from 20 days to 5 years (mean, 9 months), with 7 patients (43.75%) having an evolution of less than 3 months. Interestingly, a maxillary OOC manifesting with odontogenic sinusitis was initially considered clinically and radiologically to be a periapical inflammatory cyst.

Enucleation, with or without curettage, was performed in 13 of the OOC cases; 1 case required marsupialization followed by enucleation, and the remaining 2 cases underwent a peripheral ostectomy because of their relatively large size. Related follow-up data were available for all patients. The follow-up period ranged from 12 to 120 months, average 31.12 months, during which only one case (1/16; 6.25%) recurred. We describe this case in detail below.

The only recurrence in our series was a teenage male (13 years old) who presented with a left mandibular radiolucent lesion measuring 30 mm, which displaced the roots of neighbouring molars without paresthesia, pain, or tooth mobility. Initially, an intra-oral double exodontia with cystectomy was performed. Histological examination revealed an odontogenic cyst lined by a stratified epithelium showing orthokeratotic keratinization on its surface as well as the presence of an evident layer of granular cells (Fig. 2a-c). Eight months after the cystectomy, the patient reported a painful swelling at the site of the operation. Panoramic radiography showed an increased radiolucent mandibular area with a poorly defined lower edge; new and wider curettage of the lesion was performed. Histological examination of the recurrent lesion revealed an OOC with features of verrucous hyperplasia. Distinct endophytic and “pushing” invasion together with a well-differentiated pattern of verrucous appearance showing dysplastic features with mild cellular atypia was observed, although there was no rupture of the basement membrane. Proliferative cellular activity (positive for Ki67) was not restricted to basal cells, and p53 expression was observed in all cell nuclei in the immunohistochemical analysis (Fig. 3a-f). These morphological changes were interpreted as early signs of malignization, although there was no evidence of invasive growth based on basement membrane integrity. After surgery, clinical and radiological long-term follow-up of the patient was planned. After 2 years of quarterly visits, there was no evidence of growth or new recurrence.

In total, 38 FFPE blocks from the 16 OOC cases (including the recurrent case 13) were used for DNA extraction. Two of the 16 OOCs (cases 5 and 7) were invalid for the Anyplex II HPV28 test due to lack of beta-globin amplification and were excluded from the study (see Table 1). For the remaining 14 cases, we performed HPV analysis in 32 FFPE blocks with good DNA quality, but observed no HPV positivity in any of the primary OOCs or the recurrent lesion.