Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized Intergroup study 0099.

PURPOSE: The Southwest Oncology Group (SWOG) coordinated an Intergroup study with the participation of Radiation Therapy Oncology Group (RTOG), and Eastern Cooperative Oncology Group (ECOG). This randomized phase III trial compared chemoradiotherapy versus radiotherapy alone in patients with nasopharyngeal cancers. MATERIALS AND METHODS: Radiotherapy was administered in both arms: 1.8- to 2.0-Gy/d fractions Monday to Friday for 35 to 39 fractions for a total dose of 70 Gy. The investigational arm received chemotherapy with cisplatin 100 mg/m2 on days 1, 22, and 43 during radiotherapy; postradiotherapy, chemotherapy with cisplatin 80 mg/m2 on day 1 and fluorouracil 1,000 mg/m2/d on days 1 to 4 was administered every 4 weeks for three courses. Patients were stratified by tumor stage, nodal stage, performance status, and histology. RESULTS: Of 193 patients registered, 147 (69 radiotherapy and 78 chemoradiotherapy) were eligible for primary analysis for survival and toxicity. The median progression-free survival (PFS) time was 15 months for eligible patients on the radiotherapy arm and was not reached for the chemo-radiotherapy group. The 3-year PFS rate was 24% versus 69%, respectively (P < .001). The median survival time was 34 months for the radiotherapy group and not reached for the chemo-radiotherapy group, and the 3-year survival rate was 47% versus 78%, respectively (P = .005). One hundred eighty-five patients were included in a secondary analysis for survival. The 3-year survival rate for patients randomized to radiotherapy was 46%, and for the chemoradiotherapy group was 76% (P < .001). CONCLUSION: We conclude that chemoradiotherapy is superior to radiotherapy alone for patients with advanced nasopharyngeal cancers with respect to PFS and overall survival.

Clinicopathological determinants of positron emission tomography computed tomography fluorodeoxyglucose standardised uptake value in head and neck carcinoma.

BACKGROUND: Although positron emission tomography computed tomography has proven diagnostic and staging value in head and neck carcinoma, it does not have optimal sensitivity or specificity. The positron emission tomography computed tomography fluorodeoxyglucose standardised uptake value has been shown to be associated with carcinoma stage. This study evaluated the impact of major clinicopathological factors on the standardised uptake value at the primary site and at neck lymph node metastases. SUBJECTS AND METHODS: Two hundred and forty-three oral cavity and laryngopharyngeal carcinoma patients who underwent positron emission tomography computed tomography were included. Correlation between the positron emission tomography computed tomography standardised uptake value and various clinicopathological factors was analysed. RESULTS: A positive correlation was found between the standardised uptake value and the size and depth of tumour infiltration, and lymph node positivity. Higher standardised uptake values were seen for more advanced tumour stages. The presence of perineural invasion, lymphatic invasion and extracapsular spread were all associated with increased standardised uptake values. CONCLUSION: Most of the clinicopathological features of head and neck carcinoma which are well known to be poor prognostic factors have a significant impact on positron emission tomography computed tomography fluorodeoxyglucose standardised uptake value.

DNA copy number gains in head and neck squamous cell carcinoma.

Gene amplification, a common mechanism for oncogene activation in cancer, has been used as a tag for the identification of novel oncogenes. DNA amplification is frequently observed in head and neck squamous cell carcinoma (HNSCC) and potential oncogenes have already been reported. We applied restriction landmark genome scanning (RLGS) to study gene amplifications and low-level copy number changes in HNSCC in order to locate previously uncharacterized regions with copy number gains in primary tumor samples. A total of 63 enhanced RLGS fragments, indicative of DNA copy number changes, including gains of single alleles, were scored. Enhanced sequences were identified from 33 different chromosomal regions including those previously reported (e.g. 3q26.3 and 11q13.3) as well as novel regions (e.g. 3q29, 8q13.1, 8q22.3, 9q32, 10q24.32, 14q32.32, 17q25.1 and 20q13.33). Furthermore, our data suggest that amplicons 11q13.3 and 3q26.3-q29 may be divided into possibly two and three independent amplicons, respectively, an observation supported by published microarray expression data.

Differential targets of CpG island hypermethylation in primary and metastatic head and neck squamous cell carcinoma (HNSCC).

Head and neck squamous cell carcinomas (HNSCC) often metastasise to the cervical lymph nodes. It is known for HNSCC as well as other cancers that progression from normal tissue to primary tumour and finally to metastatic tumour is characterised by an accumulation of genetic mutations. DNA methylation, an epigenetic modification, can result in loss of gene function in cancer, similar to genetic mutations such as deletions and point mutations. We have investigated the DNA methylation phenotypes of both primary HNSCC and metastatic tumours from 13 patients using restriction landmark genomic scanning (RLGS). With this technique, we were able to assess the methylation status of an average of nearly 1300 CpG islands for each tumour. We observed that the number of CpG islands hypermethylated in metastatic tumours is significantly greater than what is found in the primary tumours overall, but not in every patient. Interestingly, the data also clearly show that many loci methylated in a patient's primary tumour are no longer methylated in the metastatic tumour of the same patient. Thus, even though metastatic HNSCC methylate a greater proportion of CpG islands than do the primary tumours, they do so at different subsets of loci. These data show an unanticipated variability in the methylation state of loci in primary and metastatic HNSCCs within the same patient. We discuss two possible explanations for how different epigenetic events might arise between the primary tumour and the metastatic tumour of a person.

Mutational status of overexpressed p16 in head and neck cancer: evidence for germline mutation of p16/p14ARF.

Inactivation of the p16 tumor suppressor gene is a common phenomenon in squamous cell carcinoma of the head and neck (SCCHN). Less commonly described is the observation of p16 overexpression in SCCHN. Since overexpression of p16 is a potent predictor of outcome in other cancers, we were interested in determining the level of expression of p16 in our SCCHN specimens as a prerequisite to later prognostic studies. We were also interested in determining the mutational status of p16 in these tumors, in order to determine whether the combination of overexpression and gene alteration may predict a different clinical outcome from overexpression alone. A total of 84 specimens of SCCHN were selected for study. These specimens were obtained from all major sites within the oral cavity, oropharynx, pharynx and larynx. The level of expression of p16 in SCCHN specimens was measured by semi-quantitative RT-PCR. In 35 cases, RNA was also isolated from matched normal tissue obtained from a negative tumor margin. In the other 49 cases, the expression level was compared with the level of expression measured in pooled normal RNA obtained from 10 specimens of normal epithelial tissue. Overexpression of p16 was documented when the level of expression in the tumor specimen was 2-fold or greater above the level of expression found in normal tissue. A total of 46 specimens demonstrated overexpression of p16 (55%). All specimens demonstrating overexpression were then subject to sequence analysis. Thirty specimens (65%) showed p16-specific gene alterations, ranging from intragenic deletions to single point mutations, and 15 of these cases concomitantly affect p14ARF. A single specimen demonstrated a silent point mutation within the p16 reading frame. This mutation produces a stop codon at residue 85 in the context of the p14ARF reading frame, predicting premature termination of p14ARF within a previously determined nucleolar localization signal. This observation suggests that in some cases at least, p14ARF may be a selective target for alteration, independently of p16. Analysis of a normal tissue specimen obtained from a negative tumor margin, and a blood sample obtained approximately five years after surgery indicate that this p14ARF-specific alteration may represent a germline mutation.

Evaluation of gemcitabine in patients with recurrent or metastatic squamous cell carcinoma of the head and neck: a Southwest Oncology Group phase II study.

A phase II trial of gemcitabine (Gemzar), a nucleoside analogue with broad activity in solid tumors, was performed in patients with recurrent or metastatic squamous cell carcinoma of the head and neck. A total of 26 eligible patients were registered to receive a dose of 1250 mg/m2 weekly for 3 weeks, followed by a 1 week rest. Toxicity was evaluable in 26 patients. Nausea and vomiting occured in 11 and 6 patients, repectively. Grade 3 or 4 hematologic toxicities were infrequent. Two patients developed neutropenic infections. One patient developed fatal liver failure which was thought due to progressive liver metastases or infection 14 days after a single dose of gemcitabine. There were no objective treatment responses (95% CI 0-13%), with a median survival of 6 months in this highly resistant disease population. Gemcitabine is not considered active enough as monotherapy for further evaluation in this disease population.

Effects of 10-hydroxycamptothecin, delivered from locally injectable poly(lactide-co-glycolide) microspheres, in a murine human oral squamous cell carcinoma regression model.

This study investigated whether local delivery of 10-hydroxycamptothecin provides effective inductive chemotherapy as assessed by significant tumor reduction. Established tumorigenic human oral squamous cell carcinoma cells were used for these experiments. The experimental groups were comprised of: control (blank (no drug) poly(lactide-co-glycolide) (PLGA) microspheres), intraperitoneal 10-hydroxycamptothecin delivery + blank microspheres, local bolus 10-hydroxycamptothecin + blank microspheres, and PLGA controlled-release microspheres. The 10-hydroxycamptothecin dose administered was 12 mg/kg (bolus-intraperitoneal, local) or controlled-release over 10 days. Regardless of delivery route, 10-hydroxycamptothecin significantly reduces tumor volume. However, PLGA microspheres provide significantly higher intratumor-drug concentrations (approximately 10 and 100 fold higher) relative to local bolus and intraperitoneal routes, respectively. Also, only the PLGA microspheres significantly reduced tumor weights. Camptothecin clinical applications are limited by drug inactivation at physiological pH and the need for sustained infusions. However, due to their acidic, camptothecin-stabilizing microclimate, PLGA microspheres could provide a novel delivery system for camptothecin-based induction chemotherapy.

A quantitative assay of telomerase activity.

PURPOSE: Telomerase is a ribonucleoprotein that extends telomeres at the ends of chromosome. Increased telomerase activity is associated with cellular immortality. The currently available assay for telomerase, i.e., telomeric repeat amplification protocol (TRAP), consists of 2 steps: (a) telomerase-mediated extension of an oligonucleotide primer by the enzyme-containing extracts of cells and tissues, and (b) amplification of the telomerase-extended primer products by polymerase chain reaction (PCR) and detection of the PCR products. It is generally accepted that the current TRAP assay lacks quantitative precision. The present study was to develop a quantitative telomerase assay with greater precision and sensitivity. METHODS: This new method used the primer extension method as in TRAP, plus the following modifications: (a) used a lysis buffer that yielded complete lysis of nuclei; (b) removal of PCR inhibitors by phenol/chloroform extraction after primer extension; and (c) used primers for the internal standard that were designed to reduce their competition with the telomerase products for PCR. RESULTS: The modified method showed a good correlation (r2 = 0.99, P < 0.001) between telomerase amount (expressed as total protein in cell lysate) and its activity (expressed as telomerase products). Compared to the conventional TRAP, the new method (a) was more sensitive (average of 5.5-fold in cultured cancer cells and >5.9-fold in patient tumors), (b) had a lower inter- and intra-day variability (>3fold), and (c) showed a 2 to 4-fold broader range of linearity in the standard curve. The higher assay sensitivity further enabled the use of a nonradioactive method, i.e., ethidium bromide staining of DNA, to detect the TRAP products, as opposed to the use of radioactive nucleotide and the more labor-intensive autoradiography mandated by the conventional TRAP. CONCLUSION: We report here a quantitative assay for telomerase activity in cultured human cancer cells and patient tumors.

Analysis of TGF-beta type I receptor for mutations and polymorphisms in head and neck cancers.

Transforming growth factor-beta receptor (TbetaR)-dependent signals are critical for cell growth and differentiation and are often disrupted during tumorigenesis. The entire coding region of TbetaR-I and flanking intron sequences from 30 head and neck carcinomas were examined for alterations using "Cold" SSCP and direct sequencing. No somatic point mutations were found in the TbetaR-I gene. In contrast, 14 polymorphic sequence changes were detected in TbetaR-I in 13 (43%) of the samples, including eight (27%) nucleotide alterations identified as polymorphisms in an exon-1 (GCG)(9) microsatellite repeat, a previously reported tumor susceptibility allele. A nine base pair deletion was found in 23% of the samples including five heterozygous and two homozygous deletions as well as single homozygous 12bp deletion. Additionally, six heterozygous polymorphisms in intronic sequences were determined, including one heterozygous C/A genotype at the +82 nucleotide position of the intron-5 intervening sequence (IVS), and five heterozygous G/A genotypes within intron-7 at the +24 nucleotide position. Exon-1 polymorphisms in the (GCG)(9) microsatellite region of the TbetaR-I gene and their association with head/neck cancers, suggest that development of these cancers may be a direct consequence of loss of responsiveness to TGF-beta mediated growth inhibition.

Long-term follow-up on an intensified treatment regimen for advanced resectable head and neck squamous cell carcinomas.

From February 1993 through July 1994, 37 patients with stage III-IV squamous cell carcinomas of the oral cavity, oropharynx, or hypopharynx (stage II-IV) were registered to a treatment regimen consisting of preoperative continuous infusion cisplatin (80 mg/m2/80 hours) with hyperfractionated external beam radiotherapy (9.1 Gy/7 fractions of 1.3 Gy BID), surgical resection, intraoperative radiotherapy (7.5 Gy), and postoperative radiotherapy (40 Gy) with concurrent cisplatin (100 mg/m2 x 2 courses). The objectives of the regimen were to improve patient compliance while also increasing treatment intensity. The purpose of this article is to report the local, regional (nodal), and distant disease control of these patients after an extended time at risk (median 40 months). Overall compliance (73%), local control at primary site (97%), and regional nodal control (95%) were excellent. The rate of distant metastasis was 19%. Absolute survival at 48 months was 45.9%.

Somatic INK4a-ARF locus mutations: a significant mechanism of gene inactivation in squamous cell carcinomas of the head and neck.

The INK4a-ARF locus is located on human chromosome 9p21 and is known to encode two functionally distinct tumor-suppressor genes. The p16(INK4a) (p16) tumor-suppressor gene product is a negative regulator of cyclin-dependent kinases 4 and 6, which in turn positively regulate progression of mammalian cells through the cell cycle. The p14(ARF) tumor-suppressor gene product specifically interacts with human double minute 2, leading to the subsequent stabilization of p53 and G(1) arrest. Previous investigations analyzing the p16 gene in squamous cell carcinomas of the head and neck (SCCHNs) have suggested the predominate inactivating events to be homozygous gene deletions and hypermethylation of the p16 promoter. Somatic mutational inactivation of p16 has been reported to be low (0-10%, with a combined incidence of 25 of 279, or 9%) and to play only a minor role in the development of SCCHN. The present study examined whether this particular mechanism of INK4a/ARF inactivation, specifically somatic mutation, has been underestimated in SCCHN by determining the mutational status of the p16 and p14(ARF) genes in 100 primary SCCHNs with the use of polymerase chain reaction technology and a highly sensitive, nonradioactive modification of single-stranded conformational polymorphism (SSCP) analysis termed "cold" SSCP. Exons 1alpha, 1beta, and 2 of INK4a/ARF were amplified using intron-based primers or a combination of intron- and exon-based primers. A total of 27 SCCHNs (27%) exhibited sequence alterations in this locus, 22 (22%) of which were somatic sequence alterations and five (5%) of which were a single polymorphism in codon 148. Of the 22 somatic alterations, 20 (91%) directly or indirectly involved exon 2, and two (9%) were located within exon 1alpha. No mutations were found in exon 1beta. All 22 somatic mutations would be expected to yield altered p16 proteins, but only 15 of them should affect p14(ARF) proteins. Specific somatic alterations included microdeletions or insertions (nine of 22, 41%), a microrearrangement (one of 22, 5%), and single nucleotide substitutions (12 of 22, 56%). In addition, we analyzed the functional characteristics of seven unique mutant p16 proteins identified in this study by assessing their ability to inhibit cyclin-dependent kinase 4 activity. Six of the seven mutant proteins tested exhibited reduced function compared with wild-type p16, ranging from minor decreases of function (twofold to eightfold) in four samples to total loss of function (29- to 38-fold decrease) in two other samples. Overall, somatic mutation of the INK4a/ARF tumor suppressor locus, resulting in functionally deficient p16 and possibly p14(ARF) proteins, seems to be a prevalent event in the development of SCCHN. Mol. Carcinog. 30:26-36, 2001.

Differential expression of a novel serine protease homologue in squamous cell carcinoma of the head and neck.

Differential gene expression between squamous cell carcinoma of the head and neck and matched normal tissue was studied by utilizing Representational Difference Analysis. Using this methodology, a novel gene, DESC1 was isolated. DESC1 possesses strong identity to the serine protease super-family. Comparison of DESC1 expression between primary squamous cell carcinoma and matched normal tissue shows that the level of DESC1 expression is reduced or absent in 11/12 SCC tissue specimens when compared to specimens of matched normal tissue. Tissue-specific expression studies further show that DESC1 expression can only be detected in tissues derived from the head and neck, and in skin, prostate and testes. Cell line studies demonstrate that DESC1 expression is epithelial-specific. Chromosomal localization studies indicate that DESC1 is located on the long arm of chromosome 4 at position q12-13.

Bedside tracheostomy in the intensive care unit: a prospective randomized trial comparing open surgical tracheostomy with endoscopically guided percutaneous dilational tracheotomy.

OBJECTIVES: Objectives of the study were 1) to analyze the complication incidence and resource utilization of two methods of bedside tracheostomy and 2) to define selection criteria for bedside tracheostomy. STUDY DESIGN: Prospective randomized trial in the setting of a tertiary care center at a university hospital. METHODS: One hundred sixty-four consecutive intubated patients selected for elective tracheostomy were enrolled. One hundred patients met selection criteria for bedside tracheostomy and were randomly assigned to either open surgical tracheostomy (50) or endoscopically guided percutaneous dilational tracheotomy(50). The remaining 64 patients received open surgical tracheostomies in the operating room. Main outcome measures were 1) perioperative and postoperative complication incidence and 2) resource utilization. RESULTS: Patients meeting our selection criteria for bedside tracheostomy had a significantly reduced perioperative complication rate compared with those who failed to meet these criteria, and subsequently underwent tracheostomy placement in the operating room (5% vs. 20%, P less than or equal to.01). No statistically significant difference was found in the perioperative complication incidence between the two methods of bedside tracheostomy. However, percutaneous tracheostomy placement at the bedside resulted in a significant increase in postoperative complication incidence (16% vs. 2%, P <.05) and incurred an additional patient charge of $436 per bedside procedure. CONCLUSIONS: This investigation prospectively confirms the safety of bedside tracheostomy placement in properly selected patients. Complication incidence and resource utilization are defined for two methods of bedside tracheostomy. The results of this study confirm that open surgical tracheostomy represents the standard of care in bedside tracheostomy placement by providing a more secure airway at a markedly reduced patient charge. These findings will aid in the development of protocols and pathways for surgical airway management in critically ill patients to maximize cost-effective, high-quality care.

Safety and efficacy of transcervical resection of parapharyngeal space neoplasms.

There are several surgical approaches for resection of parapharyngeal space (PPS) neoplasms. The purpose of this study was to evaluate local disease control, facial nerve injury, and need for mandibulotomy associated with resection of PPS neoplasms via the transcervical approach with submandibular gland excision. A retrospective chart review of 33 patients who underwent resection of a PPS neoplasm between October 1991 and July 2000 was performed. Of the 33 patients, 3 patients developed local recurrence after a median follow-up of 24 months. None of the patients experienced facial nerve paresis or paralysis. Three patients (9.1%) required a mandibulotomy for further exposure. This study demonstrated that the transcervical approach with submandibular gland excision for resection of PPS neoplasms provides excellent local disease control with minimal risk of facial nerve injury or need for mandibulotomy and/or tracheotomy.

Head and neck squamous cell carcinoma: measurement of plasma parathyroid hormone-related protein and serum and urine calcium concentrations.

OBJECTIVES: Parathyroid hormone-related protein (PTHrP) is expressed by squamous cell carcinomas. Our first objective was to examine the stability of PTHrP in normal human plasma. Our second objective was to determine whether plasma PTHrP could be used in patients with head and neck squamous cell carcinoma (HNSCC) as an indicator of tumor burden or relapse. STUDY DESIGN AND SETTING: Blood and urine samples from 55 HNSCC patients undergoing tumor resection at The Ohio State University were measured for plasma PTHrP (1-86) concentration, serum ionized calcium concentration, and urine calcium/creatinine ratio. RESULTS: Two of 55 HNSCC patients had detectable levels of plasma PTHrP. Serum ionized calcium concentrations and urinary calcium/creatinine ratios were within normal limits in all patients. CONCLUSIONS: Plasma PTHrP was not a valuable indicator of tumor presence or recurrence in our patient population. SIGNIFICANCE: Plasma PTHrP is not a useful marker of tumor presence or recurrence in patients with stage II to IV or recurrent HNSCC.

The influence of beam energy on the outcome of postoperative radiotherapy in head and neck cancer patients: secondary analysis of RTOG 85-03.

PURPOSE: To determine whether any difference in toxicity or efficacy occurs when head and neck cancer patients are treated postoperatively with (60)C0, 4 MV, or 6 MV photon beam. METHODS AND MATERIALS: This is a secondary analysis of the Intergroup Study 0034. Three hundred ninety-two patients were evaluable for comparison between treatment with (60)C0, 4 MV, or 6 MV photon beam. All patients had advanced but operable squamous cell carcinoma of the oral cavity, oropharynx, hypopharynx, or larynx. Patients were randomized following surgical resection to receive treatment with either postoperative irradiation alone, or postoperative irradiation plus three cycles of cisplatin and 5-fluorouracil. Patients were categorized as having either "low risk" or "high risk" treatment volumes based on whether the surgical margin was 5 mm or less, presence of extra capsular nodal extension, and/or carcinoma in situ at the surgical margins. Low-risk volumes received 50-54 Gy, and high-risk volumes were given 60 Gy. Patients were compared in regards to acute and late radiotherapy toxicities as well as overall survival and loco-regional control according to the beam energy used. RESULTS: One-hundred fifty-seven, 140, and 95 patients were treated by (60)C0, 4 MV, and 6 MV, respectively. No differences were seen in acute or late toxicity among treatment groups. Locoregional control was achieved in 75%, 79%, and 80% of patients treated with (60)C0, 4 MV, or 6 MV (p = 0.61). Patients treated with 6 MV had a higher incidence of ipsilateral neck failure as first event (13%) than patients treated by (60)C0 and 4 MV (9%). This difference was not statistically significant. CONCLUSION: No differences in outcome, acute, or late toxicity were discernible in patients with advanced head and neck cancer treated with (60)C0, 4 MV, or 6 MV. This result should be interpreted with caution as increased incidence, albeit nonsignificant, of ipsilateral neck recurrence was observed in patients treated with 6 MV and the power of the study to detect a statistically significant difference is small.

Hypothyroidism after treatment for nonthyroid head and neck cancer.

OBJECTIVES: To determine the incidence of posttreatment hypothyroidism in patients treated with surgery with or without radiotherapy for advanced-stage nonthyroid head and neck cancer and to make recommendations for its detection. DESIGN: A prospective study to assess the incidence and time frame of occurrence of hypothyroidism in patients by primary tumor site and treatment modality. Thyroid function tests were performed preoperatively, at the first postoperative visit, and then approximately every 6 months. Patients were followed up for up to 3 years. SETTING: Arthur G. James Cancer Hospital and Research Institute, Columbus, Ohio. PATIENTS: A total of 251 patients with nonthyroid head and neck cancer were originally enrolled; 198 patients with evaluable data were studied to determine the incidence of posttreatment hypothyroidism. Approximately 80% of the patients had advanced stage (III or IV) or recurrent cancer. RESULTS: The overall incidence of posttreatment hypothyroidism was 15% in 198 patients followed up for a mean of approximately 12 months. Hypothyroidism developed in 12% of patients treated with nonlaryngeal surgery and radiotherapy. The group undergoing total laryngectomy (with thyroid lobectomy) and radiotherapy had a 61% incidence of hypothyroidism. The average time to detection of hypothyroidism was 8.2 months. CONCLUSIONS: Approximately 15% of patients treated for advanced head and neck cancer with surgery and radiotherapy will develop hypothyroidism. Those treated with total laryngectomy and radiotherapy are at greatest risk.

Aberrant CpG-island methylation has non-random and tumour-type-specific patterns.

CpG islands frequently contain gene promoters or exons and are usually unmethylated in normal cells. Methylation of CpG islands is associated with delayed replication, condensed chromatin and inhibition of transcription initiation. The investigation of aberrant CpG-island methylation in human cancer has primarily taken a candidate gene approach, and has focused on less than 15 of the estimated 45,000 CpG islands in the genome. Here we report a global analysis of the methylation status of 1,184 unselected CpG islands in each of 98 primary human tumours using restriction landmark genomic scanning (RLGS). We estimate that an average of 600 CpG islands (range of 0 to 4,500) of the 45,000 in the genome were aberrantly methylated in the tumours, including early stage tumours. We identified patterns of CpG-island methylation that were shared within each tumour type, together with patterns and targets that displayed distinct tumour-type specificity. The expression of many of these genes was reactivated by experimental demethylation in cultured tumour cells. Thus, the methylation of particular subsets of CpG islands may have consequences for specific tumour types.