Primary intraosseous squamous cell carcinoma of the jaws: clinicopathologic presentation and prognostic factors.
Abstract: * Context.--Primary intraosseous squamous cell carcinoma is a rare form of squamous cell carcinoma, arising within the jaws.

Objective.--To clarify the clinicopathologic and behavioral spectrum of primary intraosseous squamous cell carcinoma.

Design.--The clinical and pathologic findings, treatment, and follow-up data of 39 cases of primary intraosseous squamous cell carcinoma were collected and analyzed.

Results.--Of the 39 patients, 26 (67%) were men and 13 (33%) were women. The age at diagnosis ranged from 24 to 82 years (mean, 54 years). The tumors occurred predominantly (84.6%) in the posterior mandible. Microscopically, all tumors demonstrated general features of a squamous cell carcinoma; 19 of which (49%) also showed features suggestive of an odontogenic origin. The overall survival rates were 69.8% at 2 years and 36.3% at 5 years.

Univariate analysis showed that tumors with or without odontogenic features, tumor grading, lymph node metastasis, and treatment modalities were significant prognostic factors for survival, but multivariate analysis showed that only histologic grading (relative risk, 4.43; P = .03) remained a significant prognostic factor. The cumulative probabilities of recurrence were 55.7% at 2 years and 75.9% at 5 years. Univariate analysis showed that lymph node metastasis and treatment modalities were significant prognostic factors for recurrence. On multivariate analysis, however, only lymph node metastasis remained a significant prognostic factor (relative risk, 2.54; P = .05).

Conclusions.--Primary intraosseous squamous cell carcinoma is a high-grade malignancy with frequent, regional lymph node metastasis and high rates of recurrence and mortality. The tumor grading and regional lymph node metastasis may serve as useful indicators for prognosis.
Article Type: Report
Subject: Metastasis (Diagnosis)
Squamous cell carcinoma (Diagnosis)
Squamous cell carcinoma (Prognosis)
Jaw diseases (Diagnosis)
Jaw diseases (Prognosis)
Pathology (Practice)
Authors: Huang, Jing-Wei
Luo, Hai-Yan
Li, Qiong
Li, Tie-Jun
Pub Date: 11/01/2009
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 College of American Pathologists ISSN: 1543-2165
Issue: Date: Nov, 2009 Source Volume: 133 Source Issue: 11
Topic: Event Code: 200 Management dynamics; 350 Product standards, safety, & recalls
Organization: Organization: World Health Organization
Accession Number: 230152093
Full Text: Primary intraosseous squamous cell carcinoma (PIOSCC) is a squamous cell carcinoma (SCC) arising within the jaws, having no initial connection with oral mucosa and, presumably, developing from residues of the odontogenic epithelium or from an odontogenic cyst or tumor. (1) According to Morrison and Deeley, (2) this tumor was first described by Loos in 1913. Wills, (3) in 1948, renamed it intraalveolar epidermoid carcinoma. Shear (4) later modified the name to primary intraalveolar epidermoid carcinoma. Both previous editions of the World Health Organization (WHO) classification of odontogenic tumors (5,6) recommended the term primary intraosseous carcinoma.In the new WHO classification, published in 2005, PIOSCC is designated to replace the old terms, and its subtypes include solid types, PIOSCC derived from keratocystic odontogenic tumors, and those derived from odontogenic cysts. (1) The diagnosis of PIOSCC can be difficult because it must be differentiated from other odontogenic carcinomas, such as malignant ameloblastoma, from SCCs arising from the overlying oral mucosa, from the primary tumors of the maxillary sinus or nasal mucosa, and from the tumors that have metastasized to the jaws from other primary sites. (7-11)

The main obstacle to obtaining a thorough understanding of the clinical, pathologic, and behavioral profiles of PIOSCC is the rarity of the disease. Among the published studies so far, most have dealt with single cases or reports of small series. Some reports include an exhaustive review of the published cases to define related clinicopathologic features. (7,12) Even though PIOSCC has been clearly defined by the WHO, the diagnostic criteria are obscure. Information about treatment modalities and patient outcome is also limited. Here, we report on 39 Chinese patients with PIOSCC and provide useful, new information on diagnostic criteria, treatment strategies, and prognostic factors.

MATERIALS AND METHODS

A total of 228 cases of SCC related to the jaws were retrieved from the pathologic files of Peking University School and Hospital of Stomatology (Beijing, China) during the period 19852006. On reviewing the histology, clinical details, and radiographic features, diagnosis of 39 PIOSCCs (solid type) was confirmed. The diagnostic criteria of inclusion in this series included the absence of ulceration of the oral mucosa, except when caused by other factors such as tooth extraction or trauma; histologically verified SCC; a negative result in the search for a distant primary tumor; and radiographic evidence of an osteolytic lesion that was entirely or mostly surrounded by the jaw bones. During the same period, only 3 cases were identified as PIOSCC derived from keratocystic odontogenic tumors, and no convincing case was diagnosed as PIOSCC derived from odontogenic cysts. Because of the limited data, these 3 cases were not included in the present study. Posterior maxillary SCCs with involvement of the sinus were excluded from this series because we could not distinguish them from primary carcinomas of the sinus. Clinical data on age, sex, anatomic site, duration, symptoms (pain, swelling, paresthesia and numbness of the lips, delayed healing of extraction sockets, trismus), radiographic findings, treatment modalities (surgery alone, surgery plus radiotherapy, surgery plus radiotherapy, and chemotherapy), and available follow-up data were recorded in each case. The location of the center of lesion in the maxilla or mandible was classified as the anterior region (between the right and left canines) or the posterior region (distal to the canines). To avoid the distortion caused by x-rays, the size of a lesion was expressed as the ratio of the largest diameter of the lesion and the width of the mandibular first molar in the panoramic radiographic films. By reviewing the pathologic slides, morphologic features of the tumors were evaluated, with particular emphasis on features that are suggestive of an odontogenic origin. Histologic grading of each tumor was evaluated and agreed on by 2 pathologists, using the WHO grading criteria. (1) The well-differentiated SCCs (grade 1) consisted of polygonal squamous cells arranged in orderly lobules and exhibiting numerous larger zones of keratinization. The moderately differentiated SCCs (grade 2) lost certain features so that their resemblance to squamous epithelium was less pronounced. Although scattered keratin pearls were seen in areas, anaplastic tumor cells with prominent nuclear pleomorphism were also present, particularly in the invading front. The poorly differentiated SCCs (grade 3) bore little resemblance to squamous epithelium and were formed by highly anaplastic, mitotically active, and less-cohesive tumor cells with foci of necrosis and only abortive, single-cell keratinization. The presence of regional lymph node metastasis was verified by histologic examination. Clinical outcomes (the time to relapse and time of survival) were evaluated by consulting individual clinical reports and pathology files. In addition, the patient or close relatives were contacted and were asked to complete a questionnaire regarding the current condition of the patient. Survival and disease-free survival rates were calculated using the Kaplan-Meier method. Potential prognostic factors were identified by univariate analysis using the log-rank test. Independent prognostic factors were evaluated with Cox proportional hazards model (forward-likelihood ratio). To arrive at a parsimonious multivariate model, covariates were selected for inclusion into the model only if they contributed significantly to the fit of the model (P < .50 by log-rank test). Differences at P < .05 were considered significant. Statistical analysis was performed using the Statistical Product and Service Solutions software (SPSS Inc, Chicago, Illinois).

RESULTS

Clinical Features

The 39 patients with PIOSCC ranged in age from 24 years to 82 years, with a mean of 54 years; 26 patients (66.7%) were 50 years or older. There were 26 men (66.7%) and 13 women (33.3%), with a male to female ratio of 2:1 (Figure 1). Thirty-seven tumors (94.9%) were found in the mandible, whereas 2 cases (5.1%) occurred in the maxilla. Of the 37 mandibular tumors, 33 (89.2%) involved the posterior mandible, and 4 (10.8%) occurred in the anterior region. The 2 maxillary tumors occurred in the anterior area. The median delay between presenting symptoms and diagnosis was 4 months (range, 0.5-36 months). The tumor size, as measured by the ratio of the tumor diameter to the width of the mandibular first molar, ranged in value from 1.9 to 9.8 (median, 4.4). Jaw swelling was the most common presenting feature in 32 patients (82.1%). A total of 28 of the patients (71.8%) also complained of pain, and 23 patients (59%) presented with nonhealing extraction sockets at the time of diagnosis. Of the 37 patients with mandibular tumors, 26 (70.3%) showed signs of lip or facial numbness. Of the 39 tumors, 13 (33.3%) caused varying degrees of trismus. Radiographically, osteolytic bone change was observed in all cases (Figure 2, A through D). The margins of the lesions were ill-defined or diffused in 29 cases (74.4%; Figure 2, A) and relatively well-defined in 10 cases (25.6%). Of these well-defined tumors, none of them showed corticated or sclerotic borders (0%; Figure 2, B). All tumors were radiolucent, with 3 cases (7.7%) also showing scattered calcification (Figure 2, C). Only 2 cases (5.1%) of the present series were entirely surrounded by the jawbones. In most cases, however, the lesion extended into the alveolar bone and caused its discontinuity. In such cases, the diameter of the alveolar bone defect was smaller than the anteroposterior diameter of the tumor (Figure 2, A through D), thus suggesting an origin in the jaws. Six mandibular tumors (16.2%) were found to be associated with an impacted wisdom tooth. Nine patients (24.3%) had a pathologic fracture of the mandible (Figure 2, C and D). In contrast to the prominent bone destruction, root resorption of the involved teeth was not evident, thus exhibiting a typical radiographic pattern--"floating teeth" (Figure 2, D).

[FIGURE 1 OMITTED]

Pathologic Findings

Microscopically, all tumors demonstrated the general features of a SCC (Figure 3, A and B). However, 19 tumors (48.7%) also showed features that are not commonly seen in ordinary SCCs (Figure 3, C and D). These features, suggestive of an odontogenic origin, included a relatively round or ovoid outline of the tumor nests in a prominent stroma (alveolar pattern, 10 cases [25.6%]; Figure 3, C), clear cell components (9 cases [23.1%]; Figure 3, D), and palisading arrangement of the peripheral basal cells of the tumor nests (3 cases [7.7%]). Prominent keratin formation was observed in 24 tumors (61.5%; Figure 3, A through D). Based on the WHO grading criteria, (1) 19 tumors (48.7%) were classified as grade 1 (Figure 3, A), and 20 cases (51.3%) as grade 2 (17 cases; 43.6%) or 3 (3 cases; 7.7%; Figure 3, B). Grade 2 and grade 3 tumors were combined for statistical analysis because of the few grade 3 tumors. Regional lymph node metastasis was identified histologically in 17 out of 35 patients who had lymphadenectomy (48.6%).

[FIGURE 2 OMITTED]

Treatment and Follow-up Data Analysis

In the present series, 2 patients (5.2%) were lost to follow-up after surgery. Thus, 37 patients (94.8%) had follow-up information. All these patients underwent surgery at the tumor site that consisted of partial or total maxillectomy or mandibulectomy (18 cases; 48.6%) and segmental resection (19 cases; 51.4%). Thirty five patients (94.6%) also underwent a neck dissection (primary, 33 cases [94.3%]; delayed, 2 cases [5.7%]). Fourteen patients (37.8%) were treated by surgery alone; 18 (48.7%) received radiotherapy (routine irradiation, 60 grays) as postoperative adjuvant treatment, and 5 patients (13.5%) received both chemotherapy (pingyangmycin, diamminedichloroplatinum, and others) and radiotherapy following surgery.

Follow-up of the 37 patients ranged from 2 months to 136 months (median, 19 months). Local recurrence developed in 22 patients (59.5%), including 7 patients (18.9%) who developed local recurrence 2 or more times. Four patients (10.8%) developed distant metastasis, including brain (n = 1), lung (n = 2), and long bone (n = 1). At the time of current analysis, 13 patients (35.1%) had died of disease. By the Kaplan-Meier analysis, the overall survival rates were 88.3% at 1 year, 69.8% at 2 years, and 36.3% at 5 years, respectively. Univariate analysis showed that the pathologic features (with or without odontogenic features; P = .03), histologic grading (P = .005), lymph node metastasis (P = .05), and treatment modalities (P = .02) were significant prognostic factors for survival (Table). Patients who underwent surgery alone appeared to have a better outcome than patients who received adjuvant radiotherapy or chemoradiotherapy (Figure 4, A and B). On multivariate analysis, however, only histologic grading (P = .03) emerged as a significant prognostic factor (Figure 4, C; Table). The cumulative probabilities of recurrence at 1, 2, and 5 years were 48.8%, 55.7%, and 75.9%, respectively. Univariate analysis showed that lymph node metastasis (P = .04) and treatment modalities (P = .04; Figure 4, B) were significant prognostic factors for recurrence (Table). On multivariate analysis, only lymph node metastasis (P = .05) remained a significant prognostic factor (Figure 4, D; Table).

[FIGURE 3 OMITTED]

COMMENT

Jawbones are the most common sites for epithelial cysts and tumors affecting the human skeleton. (1) These lesions are thought to arise from the epithelium involved in odontogenesis; hence, they are designated as odontogenic cysts and tumors.8 Primary intraosseous squamous cell carcinomas occur exclusively in the jawbones, and according to the new WHO classification,1 they should include carcinomas arising de novo (solid type), carcinomas arising from keratocystic odontogenic tumors, or odontogenic cysts. In the present study, we did not find any cases from our files that could be designated as carcinomas derived from odontogenic cysts, apart from the 3 cases in which definitive evidence of both a carcinoma and a preexisting keratocystic odontogenic tumor were identified. Therefore, the focus of the present study was to critically analyze the clinicopathologic and behavioral profiles of the solid type of PIOSCCs based on a series of 39 cases.

It is difficult to determine the total number of reported cases of PIOSCCs because of insufficient data. Thomas et al (7) reviewed the pooled data of 35 cases of PIOSCCs (equivalent to the solid type) and showed the mean age of the patients to be 52.5 years ranging from 4 to 81 years.

A male to female ratio of 2.5:1 was observed. The average age of patients in the present series was 54 years, and 66.7% of the patients were in their sixth decade or greater, which is consistent with the literature. There was also a preponderance in men in this series. In literature, a 4-year-old girl (13) and a 16-year-old adolescent boy (14) were reported to have PIOSCCs, which, unlike oral cavity carcinomas, are exceptional in the first 2 decades of life. Primary intraosseous squamous cell carcinoma is thought to originate from the odontogenic epithelium. This is supported by the location, in the present series, being predominantly in the posterior part of the mandible (84.6%), as are other benign odontogenic tumors or cysts. (8) The rarity of the posterior maxillary cases of PIOSCC (none from our series, and only 2 cases reported in literature, (15,16) to our knowledge) may reflect the difficulty in differentiating this lesion from carcinomas of the maxillary sinus origin.

The most common presenting symptoms of patients with PIOSCC were jaw swelling and pain. Sensory disturbances, such as lip or facial paresthesia and numbness, occurred in more than 70% of the mandibular cases. The diagnosis of PIOSCC may be complicated by secondary ulceration of the oral mucosa, such as caused by tooth extraction or trauma. Unlike many previous reports in which ulcer formation is a rare event in PIOSCC, (7,10-12) ulceration caused by a nonhealing extraction socket was noted at the time of diagnosis in 59% of our cases. This alarming difference certainly raises the possibility of diagnostic delay, which was probably due to the relatively lower regular dental attendance in the Chinese population or to previous teeth extraction without a panoramic radiograph or to having attributed the symptoms to operative complications. Some authors (9,11,17) speculated that the delayed diagnosis might contribute to the poor prognosis. In this series, a delay in correct diagnosis ranging from 2 weeks to 36 months (median: 4 months) was noted, but statistical analysis of the diagnostic delay ([less than or equal to] 3 months versus >3 months) failed to show any prognostic significance.

Radiographic examination is one of the most effective methods to detect PIOSCC. (18) However, PIOSCCs show great variation in size and shape and in the appearance of their borders. Kaffe et al (19) have proposed that an important feature of PIOSCC is the presence of indistinct margins without sclerotic outline. Our data were in accordance with this statement, as 74.7% of our cases were ill-defined, with 10 relatively well-defined lesions showing no corticated borders. Ideally, PIOSCC should be entirely surrounded by the bone on radiographic examination, which could help in confirming an intraosseous origin.

However, most tumors from the present series showed extensive destruction of the jawbones and infiltration of the soft tissues at the time of diagnosis. When discontinuity of the alveolar border of the tumor occurred, it was essential that the diameter of the alveolar cortical perforation was smaller than that of the underlying tumor, which may help in differentiating intraosseous carcinomas from tumors of the surface mucosa. It is believed that the advanced gingival SCCs may also cause underlying bone destruction, but its radiographic appearance is scallopshaped absorption with a dishing opening. A significant radiographic feature was the lack of root resorption in PIOSCCs. The tumors tended to grow around the teeth, suggesting that tumor invasion occurred along the path of least resistance. These features resulted in a unique radiographic pattern--floating teeth--a feature rarely seen in benign odontogenic tumors or cysts. In contrast to the rarity of PIOSCCs in previous reports, the number of cases in the present series is striking. This might be due to the unique patient population of our hospital. Being the major clinical center with a large head and neck oncology department in northern China, our hospital is more likely to have larger numbers of diagnostically or therapeutically difficult cases from the region.

The histologic features of PIOSCC are not pathognomonic and are often indistinguishable from SCCs arising from other squamous epithelium. (9,10,20) Thus, diagnosis of PIOSCC based on histologic examination alone is difficult, and cognizance must be taken of clinical and radiologic details. In the present series, general features of a SCC were noted in all tumors. However, nearly half of the tumors also exhibited features that are suggestive of an odontogenic origin. These features included an alveolar pattern of the tumor nests in a rich stroma background, with clear cell components and a palisading arrangement of the peripheral basal cells. Therefore, PIOSCC must be considered in the differential diagnosis of malignant tumors of odontogenic epithelium, including ameloblastic carcinoma, clear cell odontogenic carcinoma, central mucoepidermoid carcinoma, odontogenic ghost cell carcinoma, and a malignant variant of calcifying epithelial odontogenic tumor. Characteristics reminiscent of ameloblastoma, such as alveolar patterns and peripheral palisading of cells, may be exhibited in some PIOSCCs, but typical features of ameloblastic differentiation that would justify a diagnosis of ameloblastic carcinoma are lacking. Although some PIOSCCs may contain clear cell component, they all exhibited a keratin pearl formation. The later feature is not seen in clear cell odontogenic carcinoma. Mucin stains should be performed when central mucoepidermoid carcinoma is suspected, and PIOSCCs should not contain mucin-producing cells. Odontogenic ghost cell carcinoma displays, in addition to a malignant epithelial component, the classic features of calcifying cystic odontogenic tumor, including the presence of ghost cells and dentinoid materials, which are completely absent in PIOSCCs. Malignant variant of calcifying epithelial odontogenic tumors may show features of SCC, but they also exhibit additional diagnostic features, such as amyloid deposition and Liesegang ring calcifications. The salient histologic characteristic noted in the Shear (4) study of PIOSCC was the absence of keratinization. Evaluation of histologic features of this series as well as others, (7,21) however, do not support the concept that PIOSCCs are essentially nonkeratinizing SCCs. In fact 24 of our cases (61.5%) exhibited keratin formation. Subgrouping of PIOSCCs by the presence or absence of keratin formation alone did not appear to correlate with the survival or recurrence rates, but histologic grading of PIOSCCs, based on the WHO criteria, was a significant prognostic factor for patient survival. Interestingly, it has been noted that tumor grading is of limited value in SCCs of oral mucosa, although tumor staging or grading of the deeper invasive margins of the tumor could provide better prognostic information. (22-24) The staging of PIOSCCs is difficult. As SCCs, they cannot be classified as primary bone tumors, even if their location is intraosseous. Also, they cannot be classified according to the international classification of head and neck SCCs because all tumors would be classified T4, regardless of their size, because they are in close contact with the bone marrow. (12) The presence and number of regional lymph node metastases is the strongest predictor of survival in oral cavity SCCs. Cancer deaths in many patients with oral cavity SCCs stem from uncontrolled regional disease in cervical lymph nodes, often associated with local recurrence at the primary site. (25) In the present series of PIOSCC, univariate analyses showed regional lymph node metastases to be a significant factor affecting both recurrence and survival. By multivariate analysis, however, this parameter was the independent factor, only affecting recurrence rate.

[FIGURE 4 OMITTED]

Primary intraosseous squamous cell carcinoma is considered a highly malignant tumor that should be treated aggressively. Of the 37 patients with follow-up data in this series (14 [37.8%] of which were treated by surgery alone), 18 (48.7%) received adjuvant radiotherapy, and 5 (13.5%) received both chemotherapy and radiotherapy. Univariate analyses showed that treatment modality was a significant prognostic factor affecting both survival and recurrence. Interestingly, patients treated by surgery alone appeared to have a better prognosis than those treated with adjuvant therapies. Therefore, the multimodality therapy in the present series of PIOSCCs did not confer an improved outcome. Although it is possible that postoperative adjuvant chemoradiation therapy may not have been effective for patients with PIOSCC, the use of postoperative adjuvant therapy is usually considered by the surgeons for patients with advanced or high-grade tumors or tumors with positive margins after resection, who have a high likelihood of developing metastases and a poor outcome. Indeed, multivariate analysis of our data failed to confirm treatment modality as a significant variable associated with recurrence and survival. The resection margin status of the excised tumors might be another important factor affecting local recurrence and clinical outcome. Unfortunately, such data were only available in a few recent cases of our series and, thus, could not be included in the statistical analysis.

In conclusion, the overall prognosis for patients with PIOSCC is poor, with frequent regional lymph node metastasis and high recurrence and mortality rate. Our current results indicate that the histologic grading of the tumor and regional lymph node metastasis may be useful indicators for prognosis. The role of postoperative adjuvant radiotherapy or chemotherapy is uncertain and should be investigated further.

This work was supported by research grants from the National Nature Science Foundation of China (30625044, 30572048, and 30872900) and by the Specialized Research Fund for the Doctoral Program of Higher Education in China (20050001110).

References

(1.) Barnes L, Eveson JW, Reichart P, eds. Pathology and Genetics of Head and Neck Tumours. Lyon, France: IARC Press; 2005. World Health Organization Classification of Tumours; vol 9.

(2.) Morrison R, Deeley TJ. Intra-alveolar carcinoma of the jaw: treatment by supervoltage radiotherapy. Br J Radiol. 1964;35:321-326.

(3.) Wills RA. Pathology of Tumours. London, England: CV Mosby Co; 1948: 310-316.

(4.) Shear M. Primary intra-alveolar epidermoid carcinoma of the jaw. J Pathol. 1969;97(4):645-651.

(5.) Pindborg JJ, Kramer IRH, Torloni H, eds. Histological Typing of Odontogenic Tumours, Jaw Cysts and Allied Lesions. Geneva, Switzerland: World Health Organization; 1971.

(6.) Kramer IRH, Pindborg JJ, Shear M, eds. Histological Typing of Odontogenic Tumors. 2nd ed. Heidelberg, Germany: Springer-Verlag; 1992. World Health Organization International Histological Classification of Tumors.

(7.) Thomas G, Pandey M, Mathew A, et al. Primary intraosseous carcinoma of the jaw: pooled analysis of world literature and report of two new cases. Int J Oral Maxillofac Surg. 2001;30(4):349-355.

(8.) Reichart PA, Philipsen HP, eds. Odontogenic Tumors and Allied Lesion. London: Quintessence Publishing Co. Ltd; 2004.

(9.) McGowan RH. Primary intra-alveolar carcinoma: a difficult diagnosis. Br) Oral Surg. 1980;18(3):259-265.

(10.) Suei Y, Tanimoto K, Taguchi A, Wada T. Primary intraosseous carcinoma: review of the literature and diagnostic criteria. J Oral Maxillofac Surg. 1994;52(6): 580-583.

(11.) To EHW, Brown JS, Avery BS, Ward-Booth RP. Primary intraosseous carcinoma of the jaws: three new cases and a review of the literature. Br J Oral Maxillofac Surg. 1991;2 9(1):1 9-2 5.

(12.) Zwetyenga N, Pinsolle J, Rivel J, Majoufre-Lefebvre C, Faucher A, Pinsolle V. Primary intraosseous carcinoma of the jaws. Arch Otolaryngol Head Neck Surg. 2001;127(7):794-797.

(13.) Jones JH. Soft tissue and tumors in children: their structure, histogenesis and behavior. Proc R Soc Med. 1966;59(7):673-680.

(14.) Sirsat MV, Sampat MB, Shrikhande SS. Primary intra-alveolar squamous cell carcinoma of the mandible. Oral Surg Oral Med Oral Pathol. 1973;35(3): 366-371.

(15.) Punnya A, Kumar GS, Rekha K, Vandana R. Primary intraosseous odontogenic carcinoma with osteoid/dentinoid formation. J Oral Pathol Med. 2004; 33(2):121-124.

(16.) Chaisuparat R, Coletti D, Kolokythas A, Ord RA, Nikitakis NG. Primary intraosseous odontogenic carcinoma arising in an odontogenic cyst or de novo: a clinicopathologic study of six new cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2006;101(2):194-200.

(17.) Coonar HS. Primary intraosseous carcinoma of maxilla. Br Dent J. 1979; 147(2):47-48.

(18.) Lin YJ, Chen CH, Wang WC, Chen YK, Lin LM. Primary intraosseous carcinoma of the mandible. Dentomaxillofac Radiol. 2005;34(2):112-116.

(19.) Kaffe I, Ardekian L, Peled M, Machtey E, Laufer D. Radiological feature of primary intraosseous carcinoma of the jaws: analysis of the literature and report of a new case. Dentomaxillofac Radiol. 1998;27(4):209-214.

(20.) Anneroth G, Hansen LS. Variations in keratinizing odontogenic cysts and tumors. Oral Surg. 1982;54(5):530-545.

(21.) Elzay RP. Primary intraosseous carcinoma of the jaws: review and update of odontogenic carcinomas. Oral Surg Oral Med Oral Pathol. 1982;54(3):299 303.

(22.) Iro H, Waldfahrer F. Evaluation of the newly updated TNM classification of head and neck carcinoma with data from 3247 patients. Cancer. 1998;83(10): 2201-2207.

(23.) Odell EW, Jani P, Sherriff M, et al. The prognostic value of individual histologic grading parameters in small lingual squamous cell carcinomas: the importance of the pattern of invasion. Cancer. 1994;74(3):789-794.

(24.) Bryne M, Koppang HS, Lilleng R, Kjaerheim A. Malignancy grading of the deep invasive margins of oral squamous cell carcinoma has high prognostic value. J Pathol. 1992;166(4):375-381.

(25.) Kalnins IK, Leonard AG, Sako K, Razack MS, Shedd DP. Correlation between prognosis and degree of lymph node involvement in carcinoma of the oral cavity. Am J Surg. 1977;134(4):450-454.

Jing-Wei Huang, DDS, MDS; Hai-Yan Luo, DDS, MDS; Qiong Li, PhD; Tie-Jun Li, DDS, PhD

Accepted for publication January 12, 2009.

From the Department of Oral Pathology, Peking University School and Hospital of Stomatology (Drs Huang, Luo and T.-J. Li) and the School of Public Health, Peking University (Dr Q. Li), Beijing, China.

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Tie-Jun Li, MDS, Department of Oral Pathology, Peking University School and Hospital of Stomatology, 22 S Zhongguancun Ave, Haidian District, Beijing 100081, China (e-mail: litiejun22@vip.sina. com).
Univariate and Multivariate Analysis of Survival and Recurrence
of the Present Series (a)
                                                   Survival

                                  No. of                Univariate
Variables                         Cases    2 y, %       P

Age, y                                                  .29
  <55                               18     80.7
  [greater than or equal to] 55     19     57

Sex                                                     .85
  Men                               25     70.4
  Women                             12     70.1

Diagnostic delay, mo                                    .87
  [less than or equal to] 3         17     72.1
  >3                                20     69.4

Lower lip anesthesia (b)                                .68
  Absent                            10     59.3
  Present                           25     70.4

Lesion scope (c)                                        .62
  [less than or equal to] 4.4       17     60.2
  >4.4                              20     79

Histologic grading                                      .005
  Grade 1                           19     81.4
  Grade 2 and 3                     18     54.4

Pathologic features                                     .03
  With OF                           17     81.5
  Without OF                        20     56.6

Keratinization                                          .39
  Absent                            14     57.7
  Present                           23     77.3

Lymph nodes metastasis (d)                              .05
  Absent                            18     76.7
  Present                           17     56.1

Treatment                                               .02
  Surgery alone                     14     91.7
  Surgery + R                       18     62.1
  Surgery + R + C                   5      40

                                                   Survival

                                  No. of
Variables                         Cases    Cox P        RR

Age, y
  <55                               18
  [greater than or equal to] 55     19

Sex
  Men                               25
  Women                             12

Diagnostic delay, mo
  [less than or equal to] 3         17
  >3                                20

Lower lip anesthesia (b)
  Absent                            10
  Present                           25

Lesion scope (c)
  [less than or equal to] 4.4       17
  >4.4                              20

Histologic grading                         .03          4.43
  Grade 1                           19
  Grade 2 and 3                     18

Pathologic features
  With OF                           17
  Without OF                        20

Keratinization
  Absent                            14
  Present                           23

Lymph nodes metastasis (d)
  Absent                            18
  Present                           17

Treatment
  Surgery alone                     14
  Surgery + R                       18
  Surgery + R + C                   5

                                           Survival     Recurrence

                                  No. of
Variables                         Cases    95% CI       2 y, %

Age, y
  <55                               18                  60.7
  [greater than or equal to] 55     19                  66.3

Sex
  Men                               25                  51.2
  Women                             12                  88.6

Diagnostic delay, mo
  [less than or equal to] 3         17                  65.7
  >3                                20                  62.2

Lower lip anesthesia (b)
  Absent                            10                  70.4
  Present                           25                  58.5

Lesion scope (c)
  [less than or equal to] 4.4       17                  56.3
  >4.4                              20                  71.6

Histologic grading                         1.18-16.58
  Grade 1                           19                  45.3
  Grade 2 and 3                     18                  89.7

Pathologic features
  With OF                           17                  54.9
  Without OF                        20                  73.6

Keratinization
  Absent                            14                  73.3
  Present                           23                  59

Lymph nodes metastasis (d)
  Absent                            18                  46.7
  Present                           17                  80.6

Treatment
  Surgery alone                     14                  40.4
  Surgery + R                       18                  87
  Surgery + R + C                   5                   100

                                                  Recurrence

                                  No. of   Univariate
Variables                         Cases    P            Cox P

Age, y                                     .70
  <55                               18
  [greater than or equal to] 55     19

Sex                                        .33
  Men                               25
  Women                             12

Diagnostic delay, mo                       .66
  [less than or equal to] 3         17
  >3                                20

Lower lip anesthesia (b)                   .76
  Absent                            10
  Present                           25

Lesion scope (c)                           .75
  [less than or equal to] 4.4       17
  >4.4                              20

Histologic grading                         .06
  Grade 1                           19
  Grade 2 and 3                     18

Pathologic features                        .54
  With OF                           17
  Without OF                        20

Keratinization                             .56
  Absent                            14
  Present                           23

Lymph nodes metastasis (d)                 .04          .05
  Absent                            18
  Present                           17

Treatment                                  .03
  Surgery alone                     14
  Surgery + R                       18
  Surgery + R + C                   5

                                                  Recurrence

                                  No. of
Variables                         Cases    RR           95% CI

Age, y
  <55                               18
  [greater than or equal to] 55     19

Sex
  Men                               25
  Women                             12

Diagnostic delay, mo
  [less than or equal to] 3         17
  >3                                20

Lower lip anesthesia (b)
  Absent                            10
  Present                           25

Lesion scope (c)
  [less than or equal to] 4.4       17
  >4.4                              20

Histologic grading
  Grade 1                           19
  Grade 2 and 3                     18

Pathologic features
  With OF                           17
  Without OF                        20

Keratinization
  Absent                            14
  Present                           23

Lymph nodes metastasis (d)                 2.54         1.01-6.39
  Absent                            18
  Present                           17

Treatment
  Surgery alone                     14
  Surgery + R                       18
  Surgery + R + C                   5

Abbreviations: 95% CI, 95% confidence interval; C, chemotherapy;
Grade 1, well-differentiated squamous cell carcinoma (SCC); grade 2
and 3, moderately and poorly differentiated SCC; OF, odontogenic
features; R, radiotherapy; RR, relative risk.

(a) Only 37 cases were analyzed because 2 patients were lost to
follow-up.

(b) Two maxillary cases were not included.

(c) Lesion scope was calculated as the ratio of the largest diameter
of the lesion to the width of the mandibular first molar in the
panoramic radiographic films.

(d) Only 35 patients were subject to lymphadenectomy.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.