The Maxillofacial Center for Diagnostics & Research

This Page: Text and Commentary on Abstracts of Head & Neck Cancer Studies    

Abstracts, etc. Pertaining to Head & Neck Cancer

Epidemiology of carcinoma in situ of the upper aerodigestive tract. Bouquot J, Gnepp D. West Virginia University, Morgantown, WV, and St. Louis University, St. Louis, MO, U.S.A.

Reference: J Oral Pathol 1988; 17:433.

Epidemiological analyses of upper aerodigestive tract (UAT) carcinoma in situ are non-existent, hence only hospital records and surgical pathology resources are available for its descriptive characterization. The National Cancer Institute's Surveillance, Epidemiology and End Results investigation has, however, published valuable raw data pertaining to CIS of the UAT. This information is analyzed here and summarized for the first time, demonstrating the following: a male incidence rate which is 4 times greater than the female rate (0.8/1000,000 males versus 0.2/100,000 females); similar ages at diagnosis for both sexes (62.7 and 60.0 years of age for males and females, respectively); no significant urban/rural incidence differences; that CIS represents 2.8% of all UAT malignancies; that 2.0% of all CIS lesions in humans are found at UAT sites; and that CIS of the uterine cervix is not a good model for CIS of the UAT.

Incidence, time trends and follow-up of oral/pharyngeal carcinoma in situ, 1935-1984. Bouquot JE, Kurland LT, Weiland LH. West Virginia University, Morgantown, WV; Mayo Clinic, Rochester, MN

Presented to the International Association of Oral Pathologists in Toronto, Canada, 1986.
Reference: J Oral Pathol 1986;15:395

The dearth of epidemiologic studies relating to carcinoma in situ (CIS) of the mouth and pharynx has lead to the acceptance of clinicopathologic data for the descriptive characterization of this disease. Such data, however, cannot be said to represent any population beyond that of the particular hospital from which a study originated. Hence, our understanding of this disease is necessarily skewed to the more difficult or unusual cases. The present investigation is the first to offer a clinical characterization of CIS from a pool of patients known to represent virtually all cases of CIS and invasive carcinoma in a stable population (Rochester, Minnesota, 1935-84). The annual incidence of CIS in this population was 2.0/100,000 (3.7 for males; 0.7 for females), compared to a rate of 12.6 invasive carcinomas/100,000 (20.7 for males and 6.8 for females). Incidence of CIS increased with increasing age, to a maximum of 28.3/100,000 for males 75 years of age and older. Time trends of CIS incidence showed no significant change over 50 years. Clinical features and follow-up results will be reported.
                                          This study was supported in part by NIH grand AM30582.

Leukoplakia and carcinoma in situ synchronously associated with invasive oral and oropharyngeal carcinomas in Rochester, Minnesota.

Presented to the American Academy of Oral Pathology, May, 1986.

Fifty years of laryngeal precancers and early cancers in Rochester, Minnesota, USA. Bouquot J, Weiland L, Kurland L. West Virginia University, Morgantown, WV; Mayo Clinic Scottsdale, Scottsdale, AZ; Mayo Clinic, Rochester, MN.

Presented to the 7th International Symposium on Prevention & Detection of Cancer, in Nice, France, April, 1989.
Reference: Cancer Detect Prevent 1989; 14:132.

All laryngeal carcinomas (CAs, n = 52), carcinomas in situ (CISs, n = 7), and leukoplakias (LEUKs, n = 105) diagnosed in the rochester, MN population during 1935-1984 were identified from local and regional hospitals and clinics in order to determine annual incidence rates (3.3, 0.4 & 5.2/100,000, respectively, time trends (increasing incidence for CAs & LEUKs), relative frequency rates (17.8%, 2.4% & 18.5%, respectively, of all upper aerodigestive lesions of like diagnoses were laryngeal), prognosis (75% 10-year relative survival rate for CAs; 23.8% malignant transformation rate for LEUKs), and characteristics of early cancers & premalignancies.  Only 44.8% of LEUKs were biopsied & overall: 11.4% had CA on initial diagnosis; 3.9% had CIS or severe dysplasia; 27.6% had no dysplasia.  57% of Leuks were never treated.  Early CAs were typically glottic, small (confined to vocal cords), well differentiated, nonmetastatic, and of short duration 97 months), with symptoms including horseness, dysplasia, chronic cough & 'sore throat."  Comparison will be made with data from major cancer centers.

Multiple carcinomas of the mouth & throat; the 54-year experience of an entire U.S. community. J. Bouquot, L. Weiland, L. Kurland. West Virginia U., Morgantown, West Virginia, and Mayo Clinic, Rochester, Minnesota.

Presented to the American Academy of Oral Pathology, San Francisco, California; May, 1992.  
Reference: Oral Surg Oral Med Oral Pathol 1992; 74: 602.

Hospital-based investigations have concluded that as few as 1% and as many as 40% of mouth and throat cancer patients develop additional primary carcinomas of these mucosal sites. Such a wide variation probably reflects differing patient referral patterns of individual hospitals more than real differences in risk, but no population-based data have been available for comparison. The present study is the first to identify all cases or oral, pharyngeal and laryngeal carcinoma in a well-delineated population in order to determine the relative frequency and characteristics of multiple cancers in an unbiased patient sample. Medical records of Rochester, MN residents diagnosed with these cancers between 1935 and 1988 were retrieved from all primary-, secondary-, and tertiary-care hospitals where Rochester residents may have gone for care. Of 292 primary mucosal carcinomas of the upper aerodigestive tract (excluding lip vermilion) identified in 259 residents, 25 lesions in 22 persons were additional primaries of those sites (39 of 339 cancers if lip vermilion is included). Thus multiple primaries occurred in only 8.5% of UAT carcinoma patients, 12%, 8% & 6%, RESPECTIVELY, OF laryngeal, oral and pharyngeal carcinomas were second or third primaries. Also, 58% (n=14) were synchronously diagnosed. Clinicopathologic data will be presented.

Incidence of oral verrucous carcinoma in a U.S. community with minimal smokeless tobacco consumption.  J. Bouquot, L. Weiland, L. Kurland. The Maxillofacial Center, Morgantown, West Virginia; Mayo Clinic West, Phoenix, Arizona, Mayo Clinic, Rochester, Minnesota.

Presented to the Biennial Congress of the International Association of Oral Pathology in York, England, 1996.
Reference: J Oral Pathol Med 1996; 25:271.

Objective: To determine the annual incidence rate for verrucous carcinoma of the mouth and throat. Method: A population-based incidence cohort study covering a 50 year time period was conducted in the community of Rochester, Minnesota. All upper aerodigestive tract carcinomas diagnosed in this community from 1935 to 1985 were identified through the Mayo Clinic's record-linkage system. Incidence and relative frequency rates were calculated from this data. Results: Verrucous carcinoma represented 3.3% of all oral carcinomas. The age-adjusted average annual incidence rate was 0.1/100,000 person-years (0.2 for males, 0.0 for females). The incidence rate for all intraoral carcinomas was 3.1/100,000 (4.2 for males; 2.1 for females). Among males over 64 years of age the incidence rate for verrucous carcinoma was increased to 3.2/100,000 person-years. Verrucous carcinoma was more common than intraoral melanoma, basaloid squamous carcinoma or adenosquamous carcinoma, but four times less common than intraoral carcinoma in situ. This is the first incidence data for this disease.

Intraosseous salivary tissue: examples of embryonic rests, inflammatory entrapment and mature glands in 14 human jaws. J.Bouquot, D.Gnepp, J.Hietanen. The Maxillofacial Center, Morgantown, West Virginia, Brown University, Providence, Rhode Island and University of Helsinki, Helsinki, Finland.

Presented to the annual meeting of the American Academy of Oral & Maxillofacial Pathology in Hawaii, 1999.

Problem: Hundreds of salivary neoplasms have been found to be completely enclosed within the marrow spaces of the maxilla and mandible, yet non-neoplastic salivary tissue has never been convincingly identified within marrow, either separately or adjacent to such neoplasms. This situation has forced the acceptance of an inherently illogical odontogenic origin for intramedullary salivary carcinomas and adenomas. Objective: To microscopically evaluate a large number of maxillofacial marrow samples for the presence of intramedullary salivary tissue. Study design: 5,034 maxillofacial bone samples from the Latvala Inflammatory Bone Registry were microscopically reviewed for evidence of ectopic salivary inclusions within the marrow tissues. Contributing surgeons were contacted for each identified case of intraosseous salivary tissue in order to assure that all submitted tissue was removed from within the marrow spaces rather than overlying soft tissue. Results: 3 of every 1,000 marrow samples contained embryonic rests of salivary epithelium (n=9) or relatively mature salivary glands (n=5). Two of the mature glands appear to have been embedded or implanted into marrow through inflammatory perforations of overlying cortex in persons with chronic sinusitis. We also report the chance finding of incipient odontogenic epithelial neoplasms (n = 6) and odontogenic epithelial rests (n = 84) within the fatty marrow and outside the periodontal ligament spaces, confirming that not all odontogenic neoplasms are necessarily of periodontal ligament origin. Conclusion: Salivary rests and glands are present in no fewer than 3/1,000 maxillofacial marrow samples. This provides an additional and quite logical histogenetic explanation for the presence of intraosseous salivary neoplasms.

Oral cancers with leukoplakia.  Bouquot JE. The Maxillofacial Center for Diagnostics & Research, Morgantown, WV.

Invited commentary for Oral Diseases, 1999, in press.

From time to time it is good to revisit an old topic. It gives us an opportunity to refresh our memories and may actually provide new insights which a more limited antecedent understanding could not allow. The paper by Schepman et al¹ addresses once again the question of the relationship between oral leukoplakia and oral cancer, a question first voiced in print by Sir James Paget² who wondered in 1851 about the cancer-producing potential of pipe smoker's palate or "leukokeratosis."

These investigators have taken a group of oral cancer patients and identified, through careful clinical examination, leukoplakic lesions of adjacent and distant oral mucosa. They confirm previous reports that approximately one-third of oral cancers have adjacent leukoplakic lesions,³ and add the additional information that another 15% of patients have leukoplakia some distance from the cancer. Once again we encounter the idea of "field cancerization" and wonder what might be the common, unidentified mucosal alteration (premalignant condition?) which takes place in so many individuals who eventually develop oral carcinoma. In this regard it would be good to know how many patients had proliferative verrucous leukoplakia but we are not given this information.

The investigators are careful not to overstate their conclusions. They do not allege that juxtaposed leukoplakia was the definitive source of the cancers in their patients, but rather conclude simply that the carcinomas were "preceded by or associated with" the premalignant lesions. While we assume from animal and human research that malignancy truly arises from transformed precursor cells, such a study as this cannot properly address that particular question. Fortunately, the question is nicely addressed from a different perspective in another paper by the same authors.⁴

The study found little in the way of positive correlation between the presence of associated leukoplakia and such features as the grade and stage of the carcinoma, the gender of the patient, the specific site of cancer development, and the use of smoked tobacco. This seems at first glance profoundly divergent from our usual understanding of the relationship between this precancer and cancer, and we must be very cautious about accepting it as the new dogma.

In the first place, the authors make no claim to using an unbiased or unselected patient cohort. It was composed quite naturally of consecutive patients attending a hospital clinic. How representative they were of the rest of the Dutch or other Western populations may be perhaps inferred from the fact that the proportion presenting with cervical metastasis was more than twice that of the proportion found in epidemiologic investigations,⁵ although this comparison is difficult to evaluate because we are not told how many of the neck dissections in the present investigation were prophylactic in nature or proved to be negative.

Secondly, some discrepancies from previous investigations may be due to differences in study design. Previous research has looked at the leukoplakia -- oral cancer interaction from dramatically different angles, using different definitions of leukoplakia, different levels of clinical detail and different comparison or statistical techniques.

And thirdly, the discrepancies represent very real differences between the Dutch oral cancer/precancer experience and those of patients from other cultures. The heartfelt frustration is, of course, that we truly have no way of knowing which features best characterize the "classic" leukoplakic lesion. All we can do is create a composite portrait from the varied investigations now in print.

Having said all that, the paper by Schepman et al raises several interesting points in addition to those previously mentioned. Most significantly, oral floor carcinoma seems to have no strong correlation with leukoplakia, and the follow-up study by these investigators has similarly concluded that leukoplakia of this site, i.e. sublingual keratosis, does not carry an elevated risk of malignant transformation relative to other oral mucosal sites.⁴ Perhaps it is time to rethink that facet of the precancer?

Also of significance, there was no correlation between leukoplakia presence and TNM staging. My own studies have shown that oral cancers of greater size, with metastatic deposits and with poor histopathologic differentiation are less likely to be associated with leukoplakia.³ These parameters are, however, somewhat different when taken individually than when collectively analyzed as TNM stages. It is probably significant, in this light, that no TNM classification scheme has ever taken the presence or absence of adjacent leukoplakia (or any other precancer) into account. Such systems use very crude factors in their prognostic analysis. They are relatively successful in doing so but we are not yet ready to add additional precancer features, as the present investigation confirms.

And finally, it is interesting to note the lack of correlation in this study between the presence of leukoplakia and a history of tobacco smoking. Leukoplakia has been so strongly linked to the use of tobacco that this conclusion surprises me. We don't, however, know how many leukoplakias have been destroyed by the malignant processes and it may be that a significant number were thus affected in a study with more than half of its patients classified as Stage III or Stage IV cases. When these authors followed a large number of oral leukoplakia patients for more than two years, women who did not smoke were among the high-risk group, a negative correlation which has been noted by others.^4,6 The significance of the lack of correlation in the present study is unclear but worth noting.

As mentioned, it is good to revisit topics, for no other reason than to be assured that major changes have not occurred unbeknownst to us all. The study by Schepman et al has given us this assurance and added a bit of new food for thought, but we desperately need to expand the scope and depth of such studies. The next investigations should enumerate specific leukoplakia subtypes and incorporate molecular biological marker data to more significantly add to the literature. It is time to move on to the next level of analysis.