The Odontogenic Keratocyst

Abstract

The odontogenic keratocyst (OKC, currently designated by the World Health Organization as a keratocystic odontogenic tumor) is a locally aggressive, cystic jaw lesion with a putative high growth potential and a propensity for recurrence. Although it is generally agreed that some features of OKCs are those of a neoplasia, notably the relatively high proliferative rate of epithelial cells, controversies over the behavior and management of OKCs still exist. This article is intended to review this intriguing entity and to summarize the findings of recent studies related to the nature of OKCs and their clinical and therapeutic implications. Recent advances in genetic and molecular research, i.e., PTCH1 mutations and involvement of the Hedgehog signaling pathway, have led to increased knowledge of OKC pathogenesis which hints at potential new treatment options, although the question of whether the OKC is a cyst or a cystic neoplasm is yet to be answered with certainty. Since some advocate a more conservative treatment for OKCs, notably marsupialization and decompression, future treatment strategies may focus on molecular approaches and eventually reduce or eliminate the need for aggressive surgeries.

Keywords

odontogenic keratocyst keratocystic odontogenic tumor Gorlin syndrome Hedgehog signaling pathway

Introduction

The term ‘odontogenic keratocyst’ (OKC) was first introduced by Philipsen over 50 years ago to describe a group of odontogenic cysts which showed a characteristic histological appearance (Philipsen, 1956). As compared with other types of odontogenic cysts, OKCs appear to have an intrinsically higher growth potential (Main, 1970; Browne, 1971; Li et al., 1993, 1994a,b). A propensity to recur following surgical treatment, a relationship to the so-called ‘Gorlin syndrome’ (also known as nevoid basal cell carcinoma syndrome), and the potential risk of neoplastic change place OKCs in a unique position within the spectrum of odontogenic lesions. It has long been suggested that OKCs should be regarded as benign neoplasms (Ahlfors et al., 1984; Shear, 2002a,b). While the first 2 WHO classifications of odontogenic lesions (Pindborg et al., 1971; Kramer et al., 1992) put OKCs into the category of developmental odontogenic cysts, the most recent edition designates the OKC as a keratocystic odontogenic tumor (Barnes et al., 2005), implying that the lesion is a benign neoplasm. This new classification and terminology have aroused heated debates over the nature of OKCs, particularly among oral and maxillofacial surgeons and pathologists. Although considerable insight into the biological profile of the OKC has been accumulated in recent years (Gomes et al., 2009a; Mendes et al., 2010), controversies over its behavior and management still exist. The question of whether the OKC is a cyst or a cystic neoplasm is yet to be answered with certainty. This article was developed to provide an overview of the paradoxical aspects of OKCs and to review the recent advances related to the discussions over the nature of OKCs and their clinical and/or therapeutic implications.

Clinico-Pathological Profile

Overall, OKCs probably account for 8-11% of odontogenic cysts. They are present in patients over a wide age range, but peak in the 2nd and 3rd decades and are more common in males. The mandible is the most common site (Browne, 1971; Luo and Li, 2009). About half of all OKCs occur at the angle and ramus of the mandible. In the maxilla, most OKCs present in the so-called ‘globulomaxillary’ and molar areas. A considerable number of OKCs are asymptomatic and hence are detected only by incidental radiographic findings. When they are symptomatic, swelling and intra-oral drainage appear to be most common. The radiographic presentation of OKCs is variable. Typical features, such as scalloped margins or a multilocular appearance (Fig. 1A), are indicative, but other odontogenic lesions may show similar radiological findings. Thus, there are few unequivocal clinical and radiographic features specific for OKCs. Definitive diagnosis still relies on histological examination.

Figure 1.

Radiograph of a mandibular OKC, involving the angle and ramus region, depicts a multilocular radiolucency associated with an impacted third molar (A). The epithelial lining of OKC shows features typical of a corrugated parakeratinized surface layer and a regular layer of columnar basal cells with nuclear palisading (B, x400).

Typical histologic features of OKCs have been well-characterized (Philipsen, 1956; Browne, 1971), including: a thin, uniform lining of stratified squamous epithelium with a tendency to detach from the underlying connective tissue capsule; a thin corrugated surface layer of parakeratin; a spinous cell layer 8 to 4 cells in thickness, often showing intracellular edema; a regular layer of columnar basal cells with nuclear palisading; a flat epithelial-fibrous tissue junction, usually devoid of epithelial rete ridges; and a relatively thin fibrous capsule that mostly lacks inflammatory cell infiltrate (Fig. 1B). The OKC is of particular interest because of its clinically more aggressive behavior and tendency to recur after surgery. The incidence of recurrence in various reported series has varied from 2.5 to 62% (Browne, 1991; Shear, 2002a). The reason for this great variation is partly due to the varied nature of the cases published. For example, some series included cysts from patients with Gorlin syndrome, and others excluded them. Other important variables include the duration of the follow-up periods and the methods of treatment used.

Although an OKC most commonly occurs as a single lesion in the jaw of an otherwise healthy person, about 4-5% of all OKC patients have multiple cysts with other features of the so-called ‘Gorlin syndrome’ (Gorlin and Goltz, 1960; Browne, 1991). Gorlin syndrome is a rare autosomal-dominant disorder that exhibits high penetrance and variable expressivity. Clinical manifestations are extremely varied and include basal cell carcinoma of the skin, multiple OKCs of the jaws, palmar or plantar pits, and ectopic calcification of the falx cerebri, which are considered major criteria for diagnosis (Gorlin, 1995). Multiple OKCs are the most consistent and common anomaly in Gorlin syndrome, occurring in 65-100% of patients, often during the first or second decade of life (Woolgar et al., 1987). In addition, syndrome-associated OKCs are to be found in both jaws with equal frequency, in contrast to sporadic OKCs, which involve especially the lower jaw (Lo Muzio et al., 1999a). Among the various presentations, OKCs are often the first signs of Gorlin syndrome, frequently antedating the syndromic basal cell carcinomas, thereby allowing for earlier diagnosis (Lo Muzio et al., 1999b).

Evidence accumulated from various clinico-pathological observations suggests that the OKC should be regarded as a neoplasm, due to its local aggressiveness, a high tendency to recur, its association with Gorlin syndrome, and the occasional occurrence of malignant transformation (Ahlfors et al., 1984; Dabbs et al., 1994; Makowski et al., 2001; Shear, 2002a,b).

Epithelial Cell Proliferation and Differentiation

An important piece of evidence supporting the neoplastic nature of OKCs is the consistent detection of a higher level of cell-proliferative activity in the lining epithelium. Early histological studies had shown that mitotic figures are a prominent feature of OKC epithelium (Main, 1970; Browne, 1971). However, mitosis represents only the shortest phase of the cell cycle; thus its index may not be sensitive enough to reflect cellular activity. Further evidence of a greater epithelial activity in OKC linings was produced by Toller (1971), who estimated tritiated thymidine uptake in explants of cyst walls by autoradiography. The mean labeling index, expressed as the mean percentage of labeled cells per 1000 basal cells, was 13.0%, which was approximately 7 times greater than that for non-keratinizing jaw cysts (1.7%).

Immunocytochemical methods of assessing cell proliferation have particular advantages over other techniques because of the maintenance of cellular and tissue architecture, the relative simplicity of methodology, and the rapidity of results. Using quantitative techniques based on a combination of digital image analysis for histomorphometric measurement of basement membrane and manual counting of PCNA and Ki67 immunostained cells, Li et al. investigated the proliferative activity in the epithelial linings of various major and/or sub-types of odontogenic jaw cysts (Li et al., 1994a,b, 1995). The results demonstrated that OKC linings exhibited a significantly higher level of PCNA labeling, with a predominantly suprabasal location of positive cells in comparison with dentigerous and radicular cyst linings (Figs. 2A, 2B). A comparison of different subtypes of OKC linings indicated an increased number of cells expressing Ki67 in Gorlin-syndrome-related OKC than in those of non-syndromic OKCs. Interestingly, however, no significant difference was found between recurrent and non-recurrent lesions of non-syndromic OKCs (Li et al., 1995) (Fig. 2C). The demonstration of a similar Ki67 labeling index in the linings of recurrent and non-recurrent OKCs indicates that recurrence is not associated with a subgroup of lesions showing increased proliferation, supporting the concept that inappropriate surgery on the original cyst is the most plausible reason for recurrence (Li et al., 1995). Using a semi-automated image analysis system, Landini (2006) compared the epithelial lining architecture of OKCs with that of radicular cysts. While a significant difference in the quantitatively estimated thickness in the number of cell layers between linings of OKCs and radicular cysts was detected, there was no significant difference between Gorlin-syndrome-related OKCs and sporadic OKCs. Thus, the increased proliferation does not appear to be related to the thickness of syndrome-related OKC linings, which may suggest a rapid epithelial turnover. The heightened proliferative activity of Gorlin-syndrome-related OKCs could reflect the underlying genetic abnormalities in this group of individuals. (See later section regarding PTCH1.)

Figure 2.

Immunocytochemical staining of PCNA in OKC lining epithelium (A, 400x), showing marked epithelial cell labeling with a predominantly suprabasal distribution. Quantification of PCNA labeling (B) reveals a significantly higher number of positive cells in OKC linings than in dentigerous cysts (DC) and radicular cysts (RC). Note the predominant suprabasal location of PCNA⁺ cells in OKC linings. Histogram (C) showing differences of Ki67-positive cell counts among non-recurrent, recurrent, and Gorlin-syndrome-associated OKC epithelium. [The quantitative data are expressed as positive cell counts per millimeter of basement membrane, and are summarized from Li et al. (1994b, 1995).]

The consistently higher level of PCNA and Ki67 labeling in the epithelial linings of OKCs supports the hypothesis that active cell division of the lining epithelium or mural growth is more important in the pathogenesis of OKCs than in other types of odontogenic cysts. The characteristic suprabasal location of the proliferating cells predominant in OKC linings, in contrast to that in dentigerous and radicular cysts, suggests that a unique cellular proliferation and/or differentiation process occurs within this cyst type. Histologically, the columnar structure and the apparent reverse polarity of basal cells of OKC epithelium resemble the pre-ameloblasts in developing enamel organ, suggesting that they may, to some extent, undergo ameloblastic differentiation. Studies involving the transplantation of the OKC tissue walls into athymic mice have demonstrated that the characteristic features of the epithelium are retained only in the presence of its own fibrous tissue capsule (Vedtofte et al., 1982). It is therefore interesting to suggest that the unique, predominantly suprabasal distribution of proliferating cells within OKC epithelia may be a consequence of ameloblastic differentiation within the basal cell layer, due to inductive influences of underlying connective tissue (Li et al., 1994b, 1995). The nature and possible role of these inductive influences in epithelial cell proliferation and differentiation within OKCs remain to be determined.

The p53 Gene

The p53 gene product is thought to control cell growth, with its wild-type form arresting cell cycles at the G1 phase and its mutant forms promoting cell proliferation and/or malignant transformation (Lane and Benchimol, 1990; Levine et al., 1991). It is therefore interesting to determine whether abnormalities of the p53 gene are associated with the development of OKCs. Possible involvement of the p53 gene in the growth and regulation of OKCs has been suggested by immunocytochemical demonstration of p53 protein overexpression in OKC lining cells (Ogden et al., 1992; Li et al., 1994a). Quantitative study of p53 immunoreactivity demonstrated a significantly higher level of p53 labeling in OKC linings as compared with that in dentigerous and radicular cysts (Li et al., 1996), which significantly correlated with Ki67 labeling within the same series of cyst cases. Unless molecular analysis of the p53 gene is performed, however, it is unknown whether the increased p53 labeling in OKC epithelium indicates mutation of the p53 gene or overexpression of the wild-type product due to stabilization by other gene products, e.g., cdc2 protein kinase (Sturzbecher et al., 1990) or the mdm2 gene product (Wu et al., 1993). Thus, Li et al. (1996) further examined the status of the p53 gene in the immunopositive OKC cases using combined polymerase chain-reaction and single-stranded conformation polymorphism (PCR-SSCP), followed by DNA direct sequencing. The results indicated that DNA extracted from OKC, including Gorlin-syndrome-associated lesions, harbored no mutations in the p53 gene.

In other words, the epithelial lining of the OKC expresses higher levels of p53 protein than do other cyst types, which appears to correlate with cell proliferation. However, this overexpression is not due to mutation of the p53 gene, but presumably reflects overproduction and/or stabilization of normal p53 protein. As hypothesized by some authors, a high proliferation rate may result in detectable concentrations of wild-type p53 protein in cells (Mercer and Baserga, 1985; Villuendas et al., 1992). This is supported by the experimental finding of detectable levels of wild-type p53 protein in phytohemagglutinin-stimulated, rapidly proliferating lymphocytes (Mercer and Baserga, 1985) and by the occurrence of sporadic p53-positive cells in thymus and reactive lymphoid tissues (Villuendas et al., 1992). It is, therefore, reasonable to believe that the overexpression of p53 by OKC epithelium may represent a ‘feedback’ response to its high proliferative activity, rather than the cause of its intrinsic growth potential.

The PTCH1 Gene Mutation and Involvement of the Hedgehog Signaling Pathway

Perhaps the most important and intriguing evidence that appears to tip the balance of the argument over the nature of OKCs is the investigation of the PTCH1 gene. This gene is the human homologue of the Drosophila segment polarity gene, patched, which has been proven to be the disease-causing gene for the Gorlin syndrome (Hahn et al., 1996; Johnson et al., 1996). PTCH1 has been mapped to 9q22.3–31, consisting of 23 exons and encoding a transmembrane protein of 1447 amino acids with 12 transmembrane regions, 2 extracellular loops, and a putative sterol-sensing domain (Hahn et al., 1996; Johnson et al., 1996). Like neoplasms in other cancer predisposition syndromes, OKCs in Gorlin syndrome patients are multiple and appear in a random pattern; similar isolated defects are also seen occasionally in the general population. Our studies (Gu et al., 2006; Li et al., 2008; Sun et al., 2008; Pan and Li, 2009), as well as those of others (Barreto et al., 2000; Ohki et al., 2004; Song et al., 2006), have revealed that over 85% of syndromic OKCs and nearly 30% of sporadic OKCs harbored PTCH1 mutations (Table). These findings are in general agreement with those of previous studies in patients with syndromic and/or sporadic basal cell carcinomas (Lindström et al., 2006). Most of the identified frameshift or nonsense mutations lead to the synthesis of a truncated PTCH1 protein. These mutations appear to be mainly clustered into the 2 large extracellular loops of the PTCH1 protein (Fig. 3), which are important functional domains to bind Sonic hedgehog ligand. But no apparent genotype-phenotype correlation has been established (Lindström et al., 2006; Pan and Li, 2009). In addition, loss of heterozygosity (LOH) at chromosome 9q22–31, the region to which the PTCH1 gene maps, has been observed as a frequent event in syndrome-associated tumors, including OKCs (Chenevix-Trench et al., 1993; Levanat et al., 1996; Pan et al., 2010). These findings indicate that defects of PTCH1 are involved in the pathogenesis of syndromic as well as sporadic OKCs.

Table.

Summary of PTCH1 Mutations in Gorlin-syndrome-related and Sporadic OKCs (GenBank U59464.1)^a (numbering of nucleotides: +1 = A of ATG codon)

Patient	Age/Sex	Exon/Intron	Nucleotide Change	Amino Acid Definition	Functional Effect	References
GS 1	60/F	Exon 14	c.1939A>T	p.S647C	Missense	Li et al., 2008
GS 2	13/M	Exon2	c.317T>G	p.L106R	Missense	Li et al., 2008
GS 3	15/M	Exon 2	c.331delG	p.A111PfsX6	Frameshift	Li et al., 2008
GS 4	47/F	Exon16	c.2619C>A	p.Y873X	Nonsense	Li et al., 2008
		Exon 2	c.361_362insGAGC	p.L121RfsX20	Frameshift	Li et al., 2008
GS 5	21/F	Exon 16	c.2619C>A	p.Y873X	Nonsense	Gu et al., 2006
GS 6	37/M	Exon 9	c.1338_1339insGCG	p.Y446_L447insA	In-frame insertion	Gu et al., 2006
GS 7	36/F	Intron 9	c.1347+6G>A	p.V406_Q501del	Exons 9 &10 skipping, loss of 96 amino acids	Sun et al., 2008
GS 8	9/M	Exon 16	c.2619C>A	p.Y873X	Nonsense	Sun et al., 2008
		Intron 10	c.1504–1G>A	p.L450-G534del	Exons 10&11 skipping, loss of 85 amino acids	Sun et al., 2008
GS 9	44/F	Intron 15	c.2560+1G>T	p.V751DfsX49	Exon 15 skipping, frameshift	Sun et al., 2008
GS 10	14/F	Exon 6	c.863G>A	p.G288D	Missense	Sun et al., 2008
		Exon 14	c.2196_2197del	p.S733IfsX4	Frameshift	Sun et al., 2008
GS 11	43/M	Exon 9	c.1247C>G	p.T416S	Missense	Sun et al., 2008
GS 12	22/M	Exon 20	c.3440T>G	p.F1147C	Missense	Sun et al., 2008
GS 13	36/F	Exon 19	c.3244_3246dup	p.P1082dup	In-frame duplication	Sun et al., 2008
GS 14	42/M	Exon 21	c.3499G>A	p.G1167R	Missense	Sun et al., 2008
GS 15	53/M	Exon 7	c.1012C>T	p.Q338X	Nonsense	Pan and Li., 2009
GS 16	15/F	Exon 14	c.2179delT	p.C727VfsX19	Frameshift	Pan and Li., 2009
GS 17	26/F	Exon 3	c.403C>T	p.R135X	Nonsense	Pan and Li., 2009
		Exon 17	c.2824delC	p.R942GfsX20	Frameshift	Pan and Li., 2009
GS 18	NP	Exon 16	c.2760C>A	p.Y920X	Nonsense	Barreto et al., 2000
GS 19	NP	Exon 21	c.3499G>A	p.G1167R	Missense	Barreto et al., 2000
GS 20	12/M	Exon 19	c.3265T >C	p.S1089P	Missense	Song et al., 2006
GS 21	20/M	Exon 3	c.478C >T	p.Q160X	Nonsense	Song et al., 2006
GS 22	15/M	Exon 6	c.768_777delGACAAACTTC	p.N258LfsX121	Frameshift	Song et al., 2006
OKC 1	26/M	Exon 18	c.3068_3074dup	p.H1026PfsX121	Frameshift	Pan and Li., 2009
OKC 2	68/M	Exon 19	c.3300_3301insCACGTT	p.V1100_A1101insHV	In-frame insertion	Pan and Li., 2009
OKC 3	20/F	Exon 18	c.3124_3129dup	p.V1042_C1043dup	In-frame duplication	Gu et al., 2006
OKC 4	21/M	Exon 10	c.1361_1364delGTCT	p.C454X	Nonsense	Gu et al., 2006
OKC 5	48/F	Exon 23	c.3913G>T	p.D1305Y	Missense	Gu et al., 2006
OKC 6	56/M	Exon 7	c.983delA	p.H328fsX14	Frameshift	Sun et al., 2008
		Exon 9	c.1325dupT	p.A443GfsX54	Frameshift	Sun et al., 2008
OKC 7	29/M	Exon 11	c.1558_1574del	p.H520RfsX4	Frameshift	Sun et al., 2008
		Exon 16	c.2635delG	p.D879MfsX24	Frameshift	Sun et al., 2008
OKC 8	73/F	Exon 9	c.1247C>G	p.T416S	Missense	Sun et al., 2008
OKC 9	10/F	Exon 3	c.403C>T	p.R135X	Nonsense	Sun et al., 2008
OKC 10	33/F	Exon 18	c.3162dupG	p.D1055IfsX90	Frameshift	Sun et al., 2008
OKC 11	46/M	Exon 18	c.3162dupG	p.D1055IfsX90	Frameshift	Pan and Li., 2009
OKC 12	16/M	Exon 10	c.1362_1374dup	p.L455_L458dup	In-frame duplication	Pan and Li., 2009
OKC 13	NP	Exon 3	c.518_522del	p.E173AfsX77	Frameshift	Barreto et al., 2000

All the PTCH1 mutations listed are summarized from Barreto et al. (2000), Gu et al. (2006), Song et al. (2006), Li et al. (2008), Sun et al. (2008), and Pan and Li (2009). Other reports on GS patients not using OKC as tissue samples are not included in this Table. Mutational data reported by Ohki et al. (2004) are not included in this Table, due to the authors’ inclusion of several SNPs.

NP: not presented.

Figure 3.

Distribution of PTCH1 mutations (germline, black; somatic, blank) in GS-related and sporadic OKCs, reported by our group and others listed in the Table, in relation to the domain structure of the PTCH1 protein. The triangle denotes a nonsense or a frameshift mutation, the circle denotes a splice mutation at the adjacent exon junction, the square denotes a missense mutation or in-frame insertion/deletion, and the thick line denotes the sterol-sensing domain.

The aggressive behavior and higher tendency for recurrence of OKCs have been attributed to the greater proliferative activity of the epithelial linings. To clarify the role of PTCH1 mutations in OKCs, investigators have studied epithelial cell proliferation as assessed by Ki67 labeling in a total cohort of 62 OKCs (42 sporadic and 20 syndromic cases) with known PTCH1 status (Pan and Li, 2009). The epithelial Ki67 labeling in OKCs with PTCH1 mutation was significantly higher than that in cases with no PTCH1 mutation. Furthermore, OKCs harboring PTCH1 truncation-causing mutations showed an even greater Ki67 labeling index than those with non-truncation-causing mutations. These results suggest that PTCH1 mutations, particularly those causing protein truncation, are associated with OKCs showing increased proliferative activity, and thus may relate to a phenotype of higher recurrent tendency (Pan and Li, 2009).

PTCH1 is an important molecule in the so-called Hedgehog (Hh) signaling pathway. This pathway is a key regulator of embryonic development controlling cell proliferation and cell fate. As a receptor for the Sonic hedgehog protein (SHH), PTCH1 inhibits the signaling pathway by repressing the activity of Smoothened (SMO), a seven-span transmembrane protein with homology to a G-protein-coupled receptor (Stone et al., 1996) (Fig. 4A). According to this model, loss of PTCH1 function by inactivating PTCH1 mutations as well as aberrant activation of SMO by activating SMO mutations could cause constitutive, ligand-independent signal transduction that may lead to neoplastic growth (Toftgård, 2000) (Fig. 4B). In contrast to the frequent detection of PTCH1 mutations in OKCs, no pathogenic SMO mutation was detected in a total of 50 OKCs (Sun et al., 2008), among which were 16 Gorlin syndrome patients. These results suggest that mutation in SMO is extremely rare and that misregulation of the Hh signaling caused by PTCH1 inactivation is involved in the pathogenesis of OKCs (Sun et al., 2008; Pan et al., 2010).

Figure 4.

Hedgehog (Hh) signaling pathway. In the absence of ligand, the Hh signaling pathway is inactive (A). In this case, the transmembrane protein receptor PTCH1 or PTCH2 inhibits the activity of Smoothened (SMO), a seven-transmembrane protein. The transcription factor GLI, a downstream component of Hh signaling, is prevented from entering the nucleus through interactions with cytoplasmic proteins, including Suppressor of fused (SUFU). As a consequence, transcriptional activation of Hh target genes is repressed. Activation of the pathway (B) is initiated through binding of any of the 3 mammalian ligands—Sonic hedgehog (SHH shown in the Fig.), Desert hedgehog, or Indian hedgehog—to PTCH1 or PTCH2. Ligand binding results in relieving PTCH1’s inhibition of SMO, thus activating a cascade that leads to the translocation of an active form of the transcription factor GLI to the nucleus. Nuclear GLI activates the expression of target genes that are important for embryonic development, controlling cell proliferation and cell fate. The Hh pathway can be blocked at different levels (C). Several specific SMO inhibitors, such as Cyclopamine and small-molecule inhibitors Cur61414 and GDC-0449, have been identified to block activation of the Hh pathway caused by oncogenic mutations. Antibodies (Ab) directed against Sonic Hedgehog (SHH) could inhibit ligand activity. Antisense oligonucleotides targeting the transcripts of GLI may also provide novel ways of treating Hh-responsive tumors.

According to Knudson’s two-hit model of tumor suppressor genes (Knudson, 1971), two mutations—one occurring in each of the 2 alleles of the gene, or one mutation in one allele accompanied by another allelic loss of the remaining wide-type allele—are required to trigger neoplasm formation. The molecular analysis of PTCH1 in Gorlin-syndrome-associated tumors suggests that a ‘two-hit’ hypothesis is applicable to their pathogenesis (Levanat et al., 1996). These syndrome-associated tumors presumably arise from precursor cells that contain a hereditary ‘first hit’. Additional somatic mutation, LOH, or epigenetic silence of the other allele, acting as the ‘second hit’, would introduce functional inactivity of the PTCH1 protein. As a test of this hypothesis, a range of PTCH1 alteration profiles, including genetic mutation, LOH, and promoter hypermethylation, was recently investigated by Pan et al. (2010) to elucidate the possible genetic and epigenetic mechanisms of PTCH1 inactivation in syndromic and non-syndromic OKCs. Of all the 44 OKC samples (15 syndromic and 29 sporadic cases) tested, 13 cases (30%) were identified to fit the standard two-hit model, 10 of which contained 2 inactivating alleles by LOH and mutation separately, and 3 obtained disrupted alleles through 2 mutations in each of the alleles. Fourteen OKCs (32%) were found to conform to a one-hit model by LOH or mutation in a single allele, 5 of which lost a normal allele in PTCH1 locus, 7 tumors carried only one mutation, and 2 tumors harbored 2 coincident mutations in the same allele of PTCH1. After careful analysis, the remaining 17 cases (38%) failed to show any PTCH1 alteration. In addition, hypermethylation of the PTCH1 promoter was not identified in their series (Pan et al., 2010). These results suggest that numerous mechanisms are implicated in inactivation of the PTCH1 gene, in addition to the two-hit hypothesis. Perhaps the most contested exception to the two-hit hypothesis is haplo-insufficiency (absent or reduced function due to the loss or inactivation of a single allele). Evidence for the haplo-insufficiency model arises from studies of PTCH1 ^+/− mice (Wetmore et al., 2000; Zurawel et al., 2000). Mice heterozygous for PTCH1 recapitulate the typical developmental symptoms of Gorlin syndrome and develop medulloblastoma, implying that haplo-insufficiency of PTCH1 is sufficient to promote tumor formation in mice. In vivo studies have demonstrated that several mutant PTCH1 proteins could result in activation of Hh signaling through a dominant-negative mechanism, despite the production of wild-type PTCH1 (Hime et al., 2004). Thus, not only the standard two-hit model, but also the one-hit model exhibiting haplo-insufficiency or dominant-negative isoforms might be involved in the inactivation of the PTCH1 gene. The failure to detect any hit in about one-third of OKC cases examined might be explained by the presence of a multi-gene tumorigenesis model.

It is believed that tumors arise from a multi-step progression. The more advanced the tumor, the more hits it had accumulated. Our group has recently demonstrated that the percentage of two-hit cases in syndromic OKCs (53.3%) was significantly higher than that in sporadic OKCs (17.2%) (Pan et al., 2010), suggesting that more destructive hits accumulated due to decreased DNA repair efficiency and/or increased genetic instability in syndrome-related lesions. The cases with only one hit at the time might represent an early stage of tumor progression, and a second hit might eventually occur with a more severe phenotype. Thus, the evidence is mounting that genetic alterations of PTCH1 and the Hh pathway are involved in the development of OKCs. All these changes, i.e., PTCH1 mutations, LOH, two-hit mechanisms, and so on, are by definition features of a neoplastic process. However, one may still argue that these detected abnormalities are not consistent in all OKCs, but are seen in only a proportion of cases. Furthermore, many known non-neoplastic entities could harbor specific genetic abnormalities, such as SH3BP2 gene mutations in cherubism (Ueki et al., 2001) and GNAS1 gene mutations in fibrous dysplasia (Riminucci et al., 1999). All these raise the question of whether neoplasia can be defined by molecular genetic criteria.

Genetic Alterations in Other Genes

Although PTCH1 has been identified to play a confirmative role in Gorlin syndrome, other genes, such as PTCH2 (Fan et al., 2008; Xu and Li, 2008) and SUFU (Pastorino et al., 2009), might also be involved. PTCH2 encodes a putative transmembrane protein of 1203 amino acids, which has high homology to the PTCH1 protein (Smyth et al., 1999). Similar to PTCH1, PTCH2 is a receptor for SHH, and is involved in SHH/PTCH cell signaling (Carpenter et al., 1998) (Fig. 4A). Fan et al. (2008) identified a novel PTCH2 mutation in a Chinese Gorlin syndrome family, resulting in the loss of PTCH2 inhibitory function in the Hh signaling pathway. Our group also identified 2 novel PTCH2 missense mutations in 15 unrelated Gorlin syndrome patients. Interestingly, one PTCH2 mutation occurred in a patient carrying no PTCH1 mutation, but the other case carried both PTCH1 and PTCH2 mutations (Xu and Li, 2008). SUFU, encoding the human ortholog of Drosophila suppressor of fused, is a negative regulator of Hh signaling that interacts with all 3 GLI proteins (Fig. 4A) and mediates their nuclear export in the absence of SHH (Dunaeva et al., 2003). SUFU is a tumor-suppressor gene that predisposes individuals to medulloblastoma (Taylor et al., 2002). Pastorino et al. (2009) identified a SUFU germline splicing mutation in a family that was PTCH1-negative and who had signs and symptoms of Gorlin syndrome. Thus, although not as frequent as PTCH1 mutation, PTCH2 and SUFU germline mutations are detectable in a small number of Gorlin syndrome families. Whether these alterations represent clinically distinct phenotypes of Gorlin syndrome, in comparison with PTCH1 mutations, remains to be clarified with identification of additional families. Genotypic analysis performed by different research groups also demonstrated loss of heterozygosity in p16, p53, MCC, TSLC1, LTAS2, and FHIT genes (Agaram et al., 2004; Henley et al., 2005; Malcic et al., 2008). All of these genes are tumor-suppressor genes associated with different types of human neoplasia. These findings are helpful to explain the aggressive behavior of the OKC and give further support for its neoplastic nature.

Clonality Analysis

A distinguishing feature of a neoplasm is its origin from a single clone of genetically identical cells (Nowell, 1976; Secker-Walker, 1985). Thus, defining clonality provides insight into the mechanisms of cellular proliferation and growth characteristics of “proliferative” lesions. Clonality can be inferred by referencing clonal markers such as the pattern of X chromosome inactivation in lesional tissue from female patients (Allen et al., 1992). The X-linked human androgen receptor (HUMARA) gene contains a polymorphic DNA marker that reliably illustrates the pattern of X chromosome inactivation in a tissue. Normal tissues are composed of cell populations that contain random X inactivation patterns of both paternal and maternal X-chromosomes, whereas a monoclonal cell population of a neoplastic lesion contained a non-random inactivation pattern, derived either from the mother or from the father. Using this technique, Gomes et al. (2009b) investigated the clonal origin of 19 cases of odontogenic tumors, including 6 cases of OKC. Among the 16 informative cases, 12 showed a monoclonal pattern. Four of the six OKCs were monoclonal, and the remaining two were polyclonal. While emphasizing that most odontogenic tumors are clonal, including OKCs, the authors did attribute the four polyclonal cases to the contamination of samples by stromal or inflammatory cells. The two polyclonal OKCs contained a significant dispersed inflammatory cell component (Gomes et al., 2009b). Current work undertaken in the author’s laboratory also produced a rather mixed picture regarding the clonal pattern of OKCs, although the majority of OKCs appeared to be monoclonal (unpublished observations). These findings do indicate that considerable numbers of OKCs are composed of a monoclonal population of cells, lending support for their neoplastic nature.

New Treatment Strategies

The treatment of OKCs remains controversial. The challenges lie in minimizing both the risk of recurrence and the surgical morbidity. Numerous modalities, ranging from decompression alone, to simple enucleation with or without curettage, to resection, have been used in the management of OKCs (Morgan et al., 2005; Madras and Lapointe, 2008). Researchers have attempted to investigate the outcomes of these procedures systematically. However, integrating data across individual studies revealed too many inconsistencies for definitive conclusions to be drawn.

Ironically, since the tendency to treat OKCs more aggressively appeared to be reinforced by the recent WHO reclassification of OKC as a neoplasm, there have been increasing reports that OKCs can be treated for partial cure, or even complete cure, by simple marsupialization and decompression (Pogrel and Jordan, 2004; Maurette et al., 2006). The procedure involves creating a surgical window into the jaw, decompressing the intracystic high pressure, and leading to gradual reduction of the cyst size. Once again, these reports argue against the notion that OKC is a neoplasm, since typical neoplastic behavior is believed to be continued or persistent growth to occur after incomplete removal.

From a clinical point of view regarding the OKC, what really matters is not what it is called, but how it should be better managed. Recent studies related to the PTCH1 pathway have provided some insights into the development of molecular therapeutic approaches for Gorlin syndrome and its related sporadic tumors, including OKCs. The Hh pathway can be blocked at different levels, and Hh inhibitors could serve as attractive anti-tumor agents because of their specific effects on a small number of adult tissues (Pasca di Magliano and Hebrok, 2003). Several specific SMO inhibitors have been identified (Fig. 4C). Cyclopamine, a plant-based steroidal alkaloid, has been reported to inhibit cellular responses to Hh signal (Taipale et al., 2000). It specifically inhibits SMO activity by binding to its heptahelical bundle. Treatment of mice that carry Hh-dependent tumors with cyclopamine resulted in growth inhibition and regression of cancerous tissue, but did not affect the health of treated animals. Therefore, Hh inhibition causes few, if any, toxic effects on cells that do not depend on Hh signaling (Berman et al., 2002; Thayer et al., 2003). A novel small-molecule (Cur61414) inhibitor of the Hh pathway has been identified to block elevated Hh signaling activity resulting from oncogenic mutations in PTCH1. This small molecule can suppress proliferation and induce apoptosis of basaloid nests in the basal cell carcinoma model systems, but have no effect on normal skin cells (Williams et al., 2003). More recently, an orally active small molecule that targets the Hh pathway, GDC-0449, has also been reported to show anti-tumor activity in locally advanced or metastatic basal cell carcinoma in a phase 1 clinical trial (Von Hoff et al., 2009). In addition, several Hh-specific antagonists have been identified and tested. Inhibition of ligand activity has been reported with antibodies directed against SHH, which might be considered for treating tumors that are shown to require continuous ligand activity for survival (Watkins et al., 2003) (Fig. 4C). GLI can also be inhibited at the RNA level by targeting its transcripts with antisense oligonucleotides—an approach that has been used successfully in Xenopus (Pasca di Magliano and Hebrok, 2003). Zhang et al. (2006) postulated that antagonists of Hh signaling factors could effectively treat OKCs. Their suggested strategies include re-introduction of a wild-type form of PTCH1, inhibiting the SMO molecule by synthetic antagonists and suppressing the downstream transcription factors of the Hh pathway. They suggest that intracystic injection of a SMO protein-antagonist may be the most promising treatment option (Zhang et al., 2006).

Summary

The controversies over the nature of OKCs are in fact a reflection of our limited knowledge of the molecular basis of this fascinating entity. Analysis of the data accumulated so far appears to indicate that the p53 gene mutation is extremely rare in OKCs, although the level of p53 protein expression is heightened in its epithelium (Li et al., 1996; Gomes et al., 2009a). The association of OKCs with the inherited, autosomal-dominant, Gorlin syndrome appears to indicate pathogenic links or similarities in developmental pathways between isolated and syndrome-associated cases. Namely, PTCH1 mutations and misregulation of the Hh signaling pathway have a definite part to play in the development of OKCs, both in isolated cases and in syndrome patients (Li et al., 2008; Sun et al., 2008; Gomes et al., 2009a; Mendes et al., 2010; Pan et al., 2010). Despite all these recent advances, however, the evidence either supporting or refuting OKC as a cystic neoplasm is still insufficient for definitive conclusions to be drawn. The aggressive behavior and high recurrence rate of OKCs suggest a neoplastic potential and warrant special attention. Recent advances in genetic and molecular research have led to increased knowledge of OKC pathogenesis which hints at possible new treatment options. Since some investigators have advocated more conservative treatment for OKCs, namely marsupialization, future treatment strategy may even become molecular in nature and eventually reduce or eliminate the need for aggressive methods to manage the lesions. As a result of continuous efforts at understanding the molecular basis of this lesion, the various treatment protocols currently recommended will be better rationalized.

Footnotes

Acknowledgements

The author thanks Professor Tao Xu (Peking University School and Hospital of Stomatology, China) for his inspiring suggestions and recommendations. Studies performed in the author’s laboratory have been supported by research grants from the National Nature Science Foundation of China (30872900, 30625044, and 30572048). The author also thanks the following individuals for their contributions to the related research projects: Drs. Li-Sha Sun, Shuang Pan, Qing Dong, Xiao-Mei Gu, Jun-Wei Yuan, and Li-Li Xu.

References

Agaram

Collins

Barnes

Lomago

Aldeeb

Swalsky

. (2004). Molecular analysis to demonstrate that odontogenic keratocysts are neoplastic. Arch Pathol Lab Med 128:313-317.

Ahlfors

Larsson

Sjogren

(1984). The odontogenic keratocyst: a benign cystic tumor? J Oral Maxillofac Surg 42:10-19.

Allen

Zoghbi

Moseley

Rosenblatt

Belmont

(1992). Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation. Am J Hum Genet 51:1229-1239.

Barnes

Eveson

Reichart

Sidransky

, editors (2005). Pathology and genetics of head and neck tumors. World Health Organization classification of tumors. Lyon, France: IARC Press.

Barreto

Gomez

Bale

Boson

De Marco

(2000). PTCH gene mutations in odontogenic keratocysts. J Dent Res 79:1418-1422.

Berman

Karhadkar

Hallahan

Pritchard

Eberhart

Watkins

. (2002). Medulloblastoma growth inhibition by hedgehog pathway blockade. Science 297:1559-1561.

Browne

(1971). The odontogenic keratocyst. Histological features and their correlation with clinical behavior. Br Dent J 131:249-259.

Browne

, editor (1991). Investigative pathology of the odontogenic cysts. Boca Raton, FL: CRC Press.

Carpenter

Stone

Brush

Ryan

Armanini

Frantz

. (1998). Characterization of two patched receptors for the vertebrate hedgehog protein family. Proc Natl Acad Sci USA 95:13630-13634.

10.

Chenevix-Trench

Wicking

Berkman

Sharpe

Hockey

Haan

. (1993). Further localization of the gene for nevoid basal cell carcinoma syndrome (NBCCS) in 15 Australasian families: linkage and loss of heterozygosity. Am J Hum Genet 53:760-767.

11.

Dabbs

Schweitzer

Mantz

(1994). Squamous cell carcinoma arising in recurrent odontogenic keratocyst: case report and literature review. Head Neck 16:375-378.

12.

Dunaeva

Michelson

Kogerman

Toftgard

(2003). Characterization of the physical interaction of Gli proteins with SUFU proteins. J Biol Chem 278:5116-5122.

13.

Fan

Zhang

Shen

Wang

. (2008). A missense mutation in PTCH2 underlies dominantly inherited NBCCS in a Chinese family. J Med Genet 45:303-308.

14.

Gomes

Diniz

Gomez

(2009a). Review of the molecular pathogenesis of the odontogenic keratocyst. Oral Oncol 45:1011-1014.

15.

Gomes

Oliveira Cda

Castro

de Lacerda

Gomez

(2009b). Clonal nature of odontogenic tumors. J Oral Pathol Med 38:397-400.

16.

Gorlin

(1995). Nevoid basal-cell carcinoma syndrome. Dermatol Clin 13:113-125.

17.

Gorlin

Goltz

(1960). Multiple nevoid basal cell epithelioma, jaw cysts and bifid rib: a syndrome. New Engl J Med 262:908-912.

18.

Zhao

Sun

(2006). PTCH mutations in sporadic and Gorlin-syndrome-related odontogenic keratocysts. J Dent Res 85:859-863.

19.

Hahn

Wicking

Zaphiropoulous

Gailani

Shanley

Chidambaram

. (1996). Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 85:841-851.

20.

Henley

Summerlin

Tomich

Zhang

Cheng

(2005). Molecular evidence supporting the neoplastic nature of odontogenic keratocyst: a laser capture microdissection study of 15 cases. Histopathology 47:582-586.

21.

Hime

Lada

Fietz

Gillies

Passmore

Wicking

. (2004). Functional analysis in Drosophila indicates that the NBCCS/PTCH1 mutation G509V results in activation of smoothened through a dominant-negative mechanism. Dev Dyn 229:780-790.

22.

Johnson

Rothman

Xie

Goodrich

Bare

Bonifas

. (1996). Human homolog of patched, a candidate gene for the basal cell nevus syndrome. Science 272:1668-1671.

23.

Knudson

Jr (1971). Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 68:820-823.

24.

Kramer

IRH

Pindborg

Shear

, editors (1992). Histological typing of odontogenic tumors (WHO). Berlin: Springer-Verlag.

25.

Landini

(2006). Quantitative analysis of the epithelial lining architecture in radicular cysts and odontogenic keratocysts. Head Face Med 2:4.

26.

Lane

Benchimol

(1990). p53: oncogene or anti-oncogene? Genes Dev 4:1-8.

27.

Levanat

Gorlin

Fallet

Johnson

Fantasia

Bale

(1996). A two-hit model for developmental defects in Gorlin syndrome. Nat Genet 12:85-87.

28.

Levine

Momand

Finlay

(1991). The p53 tumor suppressor gene. Nature 351:453-456.

29.

Browne

Matthews

(1993). Expression of epidermal growth factor receptors by odontogenic jaw cysts. Virchows Archiv A Pathol Anat Histopathol 423:137-144.

30.

Browne

Lawrence

Matthews

(1994a). Epithelial cell proliferation in the odontogenic keratocyst. Histochemical J 26:604A.

31.

Browne

Matthews

(1994b). Quantification of PCNA⁺ cells within odontogenic jaw cyst epithelium. J Oral Pathol Med 23:184-189.

32.

Browne

Matthews

(1995). Epithelial cell proliferation in odontogenic keratocysts: a comparative immunocytochemical study of Ki67 in simple, recurrent and basal cell nevus syndrome (BCNS) associated lesions. J Oral Pathol Med 24:221-226.

33.

Browne

Prime

Paterson

Matthews

(1996). p53 expression in odontogenic keratocyst epithelium. J Oral Pathol Med 25:249-255.

34.

Yuan

Zhao

(2008). PTCH germline mutations in Chinese nevoid basal cell carcinoma syndrome patients. Oral Dis 14:174-179.

35.

Lindström

Shimokawa

Toftgård

Zaphiropoulos

(2006). PTCH mutations: distribution and analyses. Hum Mutat 27:215-219.

36.

Lo Muzio

Nocini

Savoia

Consolo

Procaccini

Zelante

. (1999a). Nevoid basal cell carcinoma syndrome. Clinical findings in 37 Italian affected individuals. Clin Genet 55:34-40.

37.

Lo Muzio

Nocini

Bucci

Pannone

Consolo

Procaccini

(1999b). Early diagnosis of nevoid basal cell carcinoma syndrome. J Am Dent Assoc 130:669-674.

38.

Luo

(2009). Odontogenic tumors: a study of 1309 cases in a Chinese population. Oral Oncol 45:706-711.

39.

Madras

Lapointe

(2008). Keratocystic odontogenic tumour: reclassification of the odontogenic keratocyst from cyst to tumour. J Can Dent Assoc 74:165-166.

40.

Main

(1970). The enlargement of epithelial jaw cysts. Odontol Revy 21:29-49.

41.

Makowski

McGuffS

Van Sickels

(2001). Squamous cell carcinoma in a maxillary odontogenic keratocyst. J Oral Maxillofac Surg 59:76-80.

42.

Malcic

Jukic

Anic

Pavelic

Kapitanovic

Kruslin

. (2008). Alterations of FHIT and P53 genes in keratocystic odontogenic tumor, dentigerous cyst and radicular cyst. J Oral Pathol Med 37:294-301.

43.

Maurette

Jorge

de Morães

(2006). Conservative treatment protocol of odontogenic keratocyst: a preliminary study. J Oral Maxillofac Surg 64:379-383.

44.

Mendes

Carvalho

van der Waal

(2010). Biological pathways involved in the aggressive behavior of the keratocystic odontogenic tumor and possible implications for molecular oriented treatment—an overview. Oral Oncol 46:19-24.

45.

Mercer

Baserga

(1985). Expression of the p53 protein during the cell cycle of human peripheral blood lymphocytes. Exp Cell Res 160:31-46.

46.

Morgan

Burton

Qian

(2005). A retrospective review of treatment of the odontogenic keratocyst. J Oral Maxillofac Surg 63:635-639.

47.

Nowell

(1976). The clonal evolution of tumor cell populations. Science 194:23-28.

48.

Ogden

Chisholm

Kiddie

Lane

(1992). p53 protein in odontogenic cysts: increased expression in some odontogenic keratocysts. J Clin Pathol 45:1007-1010.

49.

Ohki

Kumamoto

Ichinohasama

Sato

Takahashi

Ooya

(2004). PTC gene mutations and expression of SHH, PTC, SMO, and GLI-1 in odontogenic keratocysts. Int J Oral Maxillofac Surg 33:584-592.

50.

Pan

(2009). PTCH1 mutations in odontogenic keratocyst: Are they related to epithelial cell proliferation? Oral Oncol 45:861-865.

51.

Pan

Dong

Sun

(2010). Mechanisms of inactivation of PTCH1 gene in keratocystic odontogenic tumors: modification of the two-hit hypothesis. Clin Cancer Res 16:442-450.

52.

Pasca di Magliano

Hebrok

(2003). Hedgehog signaling in cancer formation and maintenance. Nat Rev Cancer 3:903-911.

53.

Pastorino

Ghiorzo

Nasti

Battistuzzi

Cusano

Marzocchi

. (2009). Identification of a SUFU germline mutation in a family with Gorlin syndrome. Am J Med Genet Part A 149:1539-1543.

54.

Philipsen

(1956). Om keratocyster (kolesteatom) i kaeberne. Tandlaegebladet 60:963-981 [in Swedish].

55.

Pindborg

Kramer

Torloni

, editors (1971). Histological typing of odontogenic tumors, jaw cysts, and allied lesions. Geneva, Switzerland: WHO.

56.

Pogrel

Jordan

(2004). Marsupialization as a definitive treatment for the odontogenic keratocyst. J Oral Maxillofac Surg 62:651-655; partial retraction in J Oral Maxillofac Surg 65:362-363, 2007.

57.

Riminucci

Liu

Corsi

Shenker

Spiegel

Robey

. (1999). The histopathology of fibrous dysplasia of bone in patients with activating mutations of the Gs alpha gene: site-specific patterns and recurrent histological hallmarks. J Pathol 187:249-258.

58.

Secker-Walker

(1985). The meaning of a clone. Cancer Genet Cytogenet 16:187-188.

59.

Shear

(2002a). The aggressive nature of the odontogenic keratocyst: Is it a benign cystic neoplasm? Part 1. Clinical and early experimental evidence of aggressive behavior. Oral Oncol 38:219-226.

60.

Shear

(2002b). The aggressive nature of the odontogenic keratocyst: Is it a benign cystic neoplasm? Part 2. Proliferation and genetic studies. Oral Oncol 38:323-331; erratum in Oral Oncol 40:107, 2004.

61.

Smyth

Narang

Evans

Heimann

Nakamura

Chenevix-Trench

. (1999). Isolation and characterization of human patched 2 (PTCH2), a putative tumor suppressor gene in basal cell carcinoma and medulloblastoma on chromosome 1p32. Hum Mol Genet 8:291-297.

62.

Song

Zhang

Peng

Wang

Bian

(2006). Germline mutations of the PTCH gene in families with odontogenic keratocysts and nevoid basal cell carcinoma syndrome. Tumor Biol 27:175-180.

63.

Stone

Hynes

Armanini

Swanson

Johnson

. (1996). The tumor-suppressor gene patched encodes a candidate receptor for Sonic hedgehog. Nature 384:129-134.

64.

Sturzbecher

Maimets

Chumakov

Brain

Addison

Simanis

. (1990). p53 interacts with p34 cdc2 in mammalian cells: implications for cell cycle control and oncogenesis. Oncogene 5:795-801.

65.

Sun

(2008). PTCH1 and SMO gene alterations in keratocystic odontogenic tumors. J Dent Res 87:575-579.

66.

Taipale

Chen

Cooper

Wang

Mann

Milenkovic

. (2000). Effects of oncogenic mutation in Smoothened and Patched can be reversed by cyclopamine. Nature 406:1005-1009.

67.

Taylor

Liu

Raffel

Hui

Mainprize

Zhang

. (2002). Mutations in SUFU predispose to medulloblastoma. Nat Genet 31:306-310.

68.

Thayer

di Magliano

Heiser

Nielsen

Roberts

Lauwers

. (2003). Hedgehog an early and late mediator of pancreatic cancer tumorigenesis. Nature 425:851-856.

69.

Toftgård

(2000). Hedgehog signaling in cancer. Cell Mol Life Sci 57:1720-1731.

70.

Toller

(1971). Autoradiography of explants from odontogenic cysts. Br Dent J 131:57-61.

71.

Ueki

Tiziani

Santanna

Fukai

Maulik

Garfinkle

. (2001). Mutations in the gene encoding c-Abl-binding protein SH3BP2 cause cherubism. Nat Genet 28:125-126.

72.

Vedtofte

Holmstrup

Dabelsteen

(1982). Human odontogenic keratocyst transplants in nude mice. Scand J Dent Res 90:306-314.

73.

Villuendas

Piris

Orradre

Mollejo

Algara

Sanchez

. (1992). p53 protein expression in lymphomas and reactive lymphoid tissue. J Pathol 166:235-241.

74.

Von Hoff

LoRusso

Rudin

Reddy

Yauch

Tibes

. (2009). Inhibition of the hedgehog pathway in advanced basal-cell carcinoma. N Engl J Med 361:1164-1172.

75.

Watkins

Berman

Burkholder

Wang

Beachy

Baylin

(2003). Hedgehog signaling within airway epithelial progenitors and in small-cell lung cancer. Nature 422:313-317.

76.

Wetmore

Eberhart

Curran

(2000). The normal patched allele is expressed in medulloblastomas from mice with heterozygous germ-line mutation of patched. Cancer Res 60:239-246.

77.

Williams

Guicherit

Zaharian

Chai

Wichterle

. (2003). Identification of a small molecule inhibitor of the Hedgehog signaling pathway: effects on basal cell carcinoma-like lesions. Proc Natl Acad Sci USA 100:4616-4621.

78.

Woolgar

Pippin

Browne

(1987). A comparative histological study of odontogenic keratocysts in basal cell naevus syndrome and control patients. J Oral Pathol 16:75-80.

79.

Bayle

Olson

Levine

(1993). The p53-mdm2 autoregulatory feedback loop. Genes Dev 7:1126-1132.

80.

(2008). PTCH2 gene alterations in keratocystic odontogenic tumors associated with nevoid basal cell carcinoma syndrome. Beijing Da Xue Xue Bao 40:15-18.

81.

Zhang

Sun

Zhao

Bian

Fan

Chen

(2006). Inhibition of SHH signaling pathway: molecular treatment strategy of odontogenic keratocyst. Med Hypotheses 67:1242-1244.

82.

Zurawel

Allen

Wechsler-Reya

Scott

Raffel

(2000). Evidence that haploinsufficiency of Ptch leads to medulloblastoma in mice. Genes Chromosomes Cancer 28:77-81.