Monosomy 3 correlates with survival but can be predicted only in patients with large epithelioid tumors. The absence of monosomy 3 is predictable only in patients who have small, spindle-cell tumors. In most patients, prediction of monosomy 3 according to tumor size and histology is unreliable.

Squamous-cell carcinoma of the head and neck (SCCHN) is the 6th most common cancer in developed countries (Parkin et al., 1988). Cigarette smoking and alcohol are major risk factors for the development of SCCHN (Elwood et al., 1984; Vokes et aZ., 1993), though other predisposing factors must also be involved since a small proportion of SCCHN patients do not have a history of tobacco or alcohol use (Wey et al., 1987).Carcinogenesis is a multi-step process which may involve interactions between oncogenes, tumour-suppressor genes (TSG) and viruses, including some specific human papillomaviruses (HPV) (Bishop, 1991;zur Hausen, 1991). There is particularly strong support for a role for specific HPV types in the pathogenesis of squamous-cell carcinoma (SCC) of the anogenital tract (Howley, 1991;zur Hausen, 1994). Other evidence also suggests a role for HPV in the development of some SCCHN, though reported HPV prevalence rates vary greatly between studies, ranging from 0-100% (Snijders et al., 1994). Part of this variability is probably due to differences in the sensitivities of the viral detection systems used, e.g., between Southern blots, dot blots and in situ hybridisation, though significant variability still exists between studies using PCR.The HPV status of cervical carcinomas appears to have prognostic significance; HPV-negative tumours follow a more aggressive course than those which are HPV-positive (Riou et aZ., 1990; Higgins et al., 1991; Crook et al., 1992). No correlation between tumour stage or histological grade has been found in oral carcinomas (Holladay and Gerald, 1993;Woods et al., 1993), but an increased prevalence of HPV has been reported in poorly differentiated laryngeal carcinomas (PCrezAyala et al., 1990). The association between HPV status of SCCHN and tumour recurrence and survival is unclear; contradictory evidence regarding risk of tumour recurrence has been reported, though the number of HPV-positive samples studied was small (Brachman et ul., 1992;Shindoh et al., 1992).The aims of this study were 2-fold: firstly, to determine and compare the prevalence of HPV in SCC arising from different anatomic sites in the head and neck; secondly, to determinc whether the presence of HPV correlatcd with specific clinicopathological parameters. HPV prevalence and type were determined in 63 SCCHN and in the adjacent normal epithelium from 30 of these tumours using PCR. HPV consensus primers, which detect a broad spectrum of mucosotropic HPV types (Snijders et aZ., 1990), and primers specific for HPV types 6, 11, 16, 18, 31 and 33 were used, followed by Southern blot hybridisation with type-specific probes. MATERIAL AND METHODS Specimen collection and patient detailsHead and neck tumour specimens were obtained over an 8-year period from patients undergoing surgery at the Royal Liverpool University Hospital (Department of Otorhinolaryngology) and Maxillofacial Unit, Walton Hospital, Liverpool. Part of the resected tissue was snap-frozen in liquid nitrogen; the remainder was fixed in formalin for subs...

To determine whether expression of the epidermal growth factor receptor (EGFR) is of prognostic value in uveal melanoma. Methods: Thirty consecutive patients treated for primary posterior uveal melanoma by enucleation or local resection were studied. Tumors were examined for EGFR and CD68 expression by immunohistochemistry on formalin-fixed, paraffin-embedded sections. Extracted DNA from paired frozen tumor and blood samples was examined for loss of heterozygosity on chromosome 3 using polymerase chain reaction-based microsatellite analysis. Immunoreactivity for EGFR was correlated with clinicopathological, chromosome 3, and follow-up data. Results: Immunoreactivity for EGFR was observed in 7 (23%) of 30 uveal melanomas, but was restricted to solitary or small groups of cells with macrophage-like morphology. Immunoreactive cells were confirmed as macrophages using an antibody to the macrophage marker CD68. Chromosome 3 loss, epithelioid cells, and microvascular loops were detected in 17 (57%), 22 (73%) and 19 (63%) of the 30 tumors, respectively. Metastatic disease was detected in 5 patients (17%). No correlation was found between any of these variables and EGFR positivity. Conclusions: The absence of EGFR immunoreactivity in tumor cells does not support the use of EGFR expression as a prognostic indicator in patients with uveal melanoma. Future EGFR studies in uveal melanoma should be interpreted with caution in view of our findings that tumor-associated macrophages can express this receptor.

The most devastating aspect of cancer is the metastasis of tumour cells to organs distant from the original tumour site. The major problem facing oncologists treating uveal melanoma, the most common cancer of the eye, is metastatic disease. To lower mortality, it is necessary to increase our understanding of the molecular genetic alterations involved in this process. Using suppression subtractive hybridisation, we have analysed differential gene expression between four primary tumours from patients who have developed clinical metastasis and four primary tumours from patients with no evidence of metastasis to date. We have identified endothelin receptor type B as differentially expressed between these tumours and confirmed this observation using comparative multiplex RT -PCR. In a further 33 tumours, reduced endothelin receptor type B expression correlated with death from metastatic disease. Reduced expression also correlated with other known prognostic indicators, including the presence of epithelioid cells, chromosome 3 allelic imbalance and chromosome 8q allelic imbalance. Endothelin receptor type B expression was also reduced in four out of four primary small cell lung carcinomas compared to normal bronchial epithelium. We also show that the observed down-regulation of endothelin receptor type B in uveal melanoma was not due to gene deletion. Our findings suggest a role for endothelin receptor type B in the metastasis of uveal melanoma and, potentially, in the metastasis of other neural crest tumours.

The relationship between human papillomavirus (HPV) type 16 infection and p53 gene mutations was investigated in squamous cell carcinomas of the head and neck (SCCHN). HPV was detected by general primer-mediated and typespecific PCR. Alterations in the p53 gene were investigated using single-strand conformation polymorphism and sequence analysis in 27 SCCHN, of which 12 were HPV 16-positive and 15 were HPV-negative. Mutations The p53 tumour-suppressor gene (TSG) plays an important role in cellular responses to DNA damage, and alteration in the function of this gene is an important factor in the development of many types of cancer (Greenblatt et al., 1994). Mutations of the p53 TSG are common in squamous cell carcinomas of the head and neck (SCCHN) and can be detected early in the pathogenesis of these tumours (Brachman et al., 1992;Boyle et al., 1993;Greenblatt et al., 1994;Brennan et al., 1995). Inactivation of p53 function, however, also may occur as a result of interactions with certain viral proteins. In particular, the E6 protein of certain high-risk human papillomaviruses (HPVs) can bind to the p53 protein, promote its degradation and disrupt p53-mediated responses to DNA damage (Scheffner et al., 1990;Kessis et al., 1993). These findings raised the possibility that the presence of high-risk HPV types might obviate a requirement for p53 gene alterations, and initial studies of cervical carcinoma did indeed support an inverse correlation between the presence of HPV and p53 mutations . Further studies, however, have suggested that this apparently negative association may reflect an overall low prevalence of p53 mutations in tumours from this site, irrespective of HPV infection (Fujita et al., 1992;Helland et al., 1993).SCCHN provide an alternative group of tumours for investigating the relationship between HPV and p53 gene alterations as over 40% of these carcinomas have been reported to contain p53 mutations (Brachman et al., 1992;Boyle et al., 1993) and high-risk HPV types, predominantly HPV 16, have been detected in over 20% of tumours (Snijders et al., 1996). The relationship between HPV infection and p53 gene alterations in head and neck cancer is uncertain. The majority of studies of SCCHN have used p53 immunodetection rather than mutational analysis (Gorgoulis et al., 1994;Salam et al., 1995) or have concentrated largely on cell lines (Brachman et al., 1992;Min et al., 1994;Yeudall et al., 1995). Brachman et al. (1992) and Min et al. (1994) did not detect HPV 16 or 18 and p53 mutations simultaneously, though the number of HPV-positive samples in each study was very small, comprising only 3 and 2 samples, respectively. In contrast, Yeudall et al. (1995) detected HPV 16 in 2 of 8 oral carcinoma cell lines containing p53 mutations, though in both cases HPV 16 was undetectable in cells in later passages. Results from a study by Snijders et al. (1994a), however, indicate that high-risk HPV types and p53 mutations co-exist in a small number of tonsillar carcinomas; p53 mutations were detected in 2/11 HPV-...

Loss of heterozygosity (LOH) had been widely used to assess genetic instability in tumours and a high LOH on chromosome arms 3p, 9p and 17p has been considered to be a common event in squamous cell carcinoma of the head and neck (SCCHN). We have investigated LOH in 52 SCCHN using a range of microsatellite markers. LOH was observed in 69% of individuals on 17p using seven markers, in 64% of individuals on 3p using 17 markers and in 61% of individuals on 9p using 11 markers. Fractional allele loss (FAL) has been calculated for each tumour (FAL is the number of chromosomal arms showing LOH divided by the number of informative chromosomal arms) and a median FAL value of 0.25 was obtained in the 52 SCCHN studied. The LOH data were examined on the basis of FAL scores: low FAL (LFAL), 0.00-0.19; medium FAL (MFAL), 0.20-0.32; high FAL (HFAL), 0.33-0.88. HFAL tumours demonstrated a significantly higher LOH on chromosome arms 3p, 9p and 17p, with 94% LOH on 3p, 94% on 9p and 100% on 17p compared with LFAL tumours. Six of the 16 patients in the LFAL group were found to have no LOH on 3p, 9p or 17p and of these four had LOH at other sites, on chromosomes 2p25-p24, 5q21-22, 7pter-p22, 8q13-q22.1, 11q23.3, 13q32, 17q, 18p11.21, 18q21.31 and 19q12-q13.1. These results indicate that LFAL patients form a subset of SCCHN tumours with distinct molecular initiating events which may represent a discrete genetic population.

No abstract