| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Letters to the Editor |
Department of Otolaryngology/Head and Neck Surgery, Vrije Universiteit Medical Center, 1007 MB Amsterdam, the Netherlands
Letter
We read the recent study of Rosin et al. in Cancer Research (1) with great interest. The authors used LOH1 analysis to assess the risk that precursor lesions in patients after being treated for their primary p.o. tumor develop into new second p.o. malignancies. Although we fully acknowledge the importance of the question addressed in this study, we would like to emphasize that a critical issue is not included in the experimental design with possible consequences for the interpretation of the results and thereby the implications of the findings.
The definition of second p.o. malignancy needs further description. It is well known that patients with p.o. cancers have a high risk for developing local recurrences even when the resection margins were found to be tumor free after thorough histopathological examination. In relation to the index tumor, a recurrent tumor is defined to be <2 cm away and occur within a time interval of 3 years, according to current clinical criteria (2
, 3)
. In addition, these patients are also at risk for developing second primary tumors (>2 cm away or >3-year interval) in the same or adjacent anatomical area. Although the authors have recognized this distinction, they refer only to "second oral malignancy," because there is no difference in the proportion of cases with LOH at 3p and/or 9p that developed into second p.o. malignancy before and after 3 years (see "Materials and Methods"). We agree with the notion that second primary tumors arise from precursor lesions after therapy, as has clearly been described in previous studies (4, 5, 6, 7, 8)
. However, and this should be stressed, we do not agree with the assumption that local recurrences exclusively originate from precursor lesions. Local recurrences might also originate from remaining tumor cells as we and others (3
, 7
, 9)
discussed in previous publications. In fact, we recently determined that 
60% of the local recurrences originate from tumor cells left behind2
in patients with histologically tumor-free resection margins. Apparently, we are dealing in significant proportion of patients with nonresected tumor cells that can only be detected with sensitive molecular methods. The percentage of tumor-derived recurrences might even be higher in patients selected on the basis of "treatment with curative intent," as was used as criterion in the present study. The authors touched on this problem by stating that they did not find differences in risk profiles in the cases that developed a second p.o. malignancy within and after 3 years. In our view, this is not sufficient. The question is whether the authors have performed a detailed molecular analysis of both the tumor and local recurrence to check whether they have included local recurrences that originate from nonresected tumor cells. This is the only way to establish the value of different LOH patterns for risk assessment in post-treatment p.o. precursor lesions. It may even be that after correction for the proportion of tumor-derived recurrences, different and even higher risk factors will be obtained.
Although we do not question the importance of the presented findings, the current study may tell only a part of the story of the development of recurrent or second p.o. malignancies.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 The abbreviation used is: LOH, loss of heterozygosity. 
2 M. P. Tabor, V. M. M. van Houten, C. R. Leemans, J. A. Kummer, D. J. Kuik, M. W. M. van den Brekel, G. B. Snow, B. J. M. Braakhuis, and R. H. Brakenhoff. Elucidation of the pathobiology of locally recurrent head and neck cancer despite seemingly radical surgery, submitted for publication.
3 To whom requests for reprints should be addressed, at Section Tumor Biology, Department of Otolaryngology/Head and Neck Surgery, Vrije Universiteit Medical Center, Room 1 D 116, P.O. Box 7057, 1007 MB. Phone: 31 20 4443690; Fax: 31 20 4440983; E-mail: bjm.braakhuis{at}vumc.nl
Received 1/15/03. Accepted 3/ 6/03.
REFERENCES
British Columbia Cancer Research Centre, 601 West 10th Avenue, Vancouver, British Columbia, V5Z 1L3 Canada, and, School of Kinesiology, Simon Fraser University, Burnaby, British Columbia, V5A 1S6 Canada
British Columbia Cancer Research Centre, 601 West 10th Avenue, Vancouver, British Columbia, V5Z 1L3 Canada
Faculty of Dentistry, University of British Columbia, 2199 Wesbrook Mall, Vancouver, British Columbia, V6T 1Z3 Canada
Reply1
The term "second oral malignancy" (SOM)2 (1) , as used in Rosin et al. (2) , refers to second tumors in the oral cavity, regardless of origin. This includes recurrent tumors developing from residual tumor cells, SPTs developing from independently altered mucosal cells, local metastasis, and/or SFTs. The latter term (second field tumor) has been recently proposed by Braakhuis et al. (3 , 4) for tumors derived from the same genetically altered mucosal field as that of the primary tumors, sharing some but not all genetic markers, and possibly representing lesions that were originally related but diverged at a later stage.
We share the view that molecular investigation of the origin of SOMs promises to improve classification, because current clinicopathological criteria have limited value in delineating the origin of SOM. However, even molecular tools, such as LOH and p53 mutation analyses, used to investigate the source of SOMs (5 6 7) may not always differentiate among these tumors with complete confidence. The complexity of clonal evolution and resulting heterogeneity of subpopulations within a tumor can complicate interpretation. SOMs developing at the same (or adjacent) anatomical site with the same molecular alterations as the primary squamous cell carcinoma are most likely recurrent tumors. Conversely, SOMs with distinctly different patterns of genetic alterations are likely to be clonally independent and hence "true" SPTs. Unfortunately, absolute concordance (or disconcordance) between a primary tumor and an SOM is uncommon, and, in most cases, primary tumors and SOMs will share some but not all alterations.
Currently no consensus exists on how the latter SOMs should be classified. Lesions that share an identical novel microsatellite shift or a signature gene-specific mutation are likely to be clonally related. However, establishing a relationship based on more common genetic events, such as LOH at microsatellite markers, can be less distinctive. One approach is to stratify markers into early and late events, and then compare these patterns in the index and SOM. Presumably, recurrence and SFT would share the same early events with the index tumor (e.g., 3p or 9p loss) but differ in late events, with recurrence more closely resembling the index tumor. However, even in recurrent tumors, the degree of similarity is affected by clonal evolution or heterogeneity. Furthermore, even greater difficulty is encountered when this approach is used to differentiate SFTs from SPTs. SFTs and index tumors are presumed to harbor the same early events, but SPTs do not. However, because such events are found in high frequency in all tumors, similarity can occur by chance alone. Extensive allelotyping with multiple markers for each chromosome arm may be necessary to improve such classification (5) . Even then, some level of disconcordance attributable to clonal evolution is expected. Classification is based on proportional similarities or dissimilarities in the markers using arbitrary (subjective) cutoffs with the presumption that true recurrence will share most markers with the index tumors, whereas SFT will share some and SPT will share few or no markers.
Because of this lack of a reliable method in differentiating among SOMs, we feel it may be premature to subclassify the SOMs in our sample set (2) until we have a better method of delineation. Comparing multiple oral samples from the same patients might be an effective approach to delineation of clonal relatedness. Jang et al. (5) established the clonal relationship by stringent criteria that involved an assessment of the probability of two lesions exhibiting the same allelotypes by chance alone. Remarkably, a sequence of genetic events on three arms was reconstructed using 33 biopsies in four patients, revealing multiple lineages of related clones. These data point to the value of repeated sampling within a high-risk field to facilitate the delineation of clones and to establish criteria for molecular subclassification of SOMs.
Meanwhile, the issue at hand is to reduce SOM through the development of tools that assist the clinician during patient follow-up. We propose broadening studies in oral cancer patients from the focus on subclassification of SOM to risk prediction through temporal analysis of clinicopathological and molecular patterns in high-risk mucosal fields. At present, despite aggressive monitoring and the awareness of the high rate of SOM, little improvement in prognosis has occurred. A major problem is that early changes leading to SOM may not be clinically apparent or, if present, not readily differentiated from reactive changes induced by treatment. In either case, a biopsy may not be available for either histological or molecular analysis. Improvement in our ability to detect tissue that requires biopsy is a pressing issue. Similarly, even when a lesion is biopsied, the risk prediction is still limited when pronounced dysplasia is absent. The Rosin et al. article (2) was intended to test whether a set of microsatellite markers previously shown to be predictive of progression for oral premalignant lesions would also predict development of SOM for leukoplakia developing at the former cancer site. (We did not set out to distinguish the two classes; the leukoplakia could contain residual tumor cells or premalignant clones developing at the former cancer site). Loss of 3p14 and 9p21 was shown in our study to be a simple tool for predicting SOM development at former cancer sites without distinction between recurrence and SPT.
We think that using a combination of clinicopathological and molecular alterations to assess the tissue left behind after cancer treatment may be a practical approach to risk assessment. To this end, we have established a prospective study of 200 oral cancer patients that are being followed using both traditional and experimental diagnostic procedures (toluidine blue, fluorescence imaging) in parallel with molecular analyses of brushings and biopsies of this field to develop a risk model for clinical management of patients. Early studies from this laboratory have already established that LOH at 3p and 9p can be identified in exfoliated cells collected from clinically normal mucosa at previous cancer sites months before the appearance of leukoplakia and the subsequent development of SOM (8 , 9) . This noninvasive approach allows the repeated monitoring of cancer sites even when the lesion is innocuous or not apparent clinically. We and others have also shown that toluidine blue staining localizes to sites containing high-risk molecular clones and, hence, can be used to target the site of exfoliated cell scrapes or biopsies (1 , 10) .
In conclusion, both subclassification and the development of cancer risk markers of SOM are critical to the improvement of prognosis for oral cancer patients. We advocate temporal analysis and echo the call of Braakhuis et al. (4) and others for longitudinal studies to establish clinical and molecular patterns in high-risk mucosal fields.
FOOTNOTES
Supported by NIH Grant R02 DE13124 from the National Institute of Dental and Craniofacial Research
The abbreviations used are: SOM, second oral malignancy; SPT, second primary tumor; SFT, second field tumor; LOH, loss of heterozygosity.
To whom requests for reprints should be addressed, at Faculty of Dentistry, the University of British Columbia, 2199 Wesbrook Mall, Vancouver, BC, V6T 1Z3 Canada. Phone: 604-877-6123; Fax: 604-877-1868; E-mail: Lewei{at}shaw.ca
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
