Oral presentations

Introduction: The incidence of oesophageal adenocarcinoma (OAC) has increased dramatically over recent years and Barrett's oesophagus (BO) is the most established risk factor for its development. Endoscopic surveillance of BO has been widely advocated but hinges on assessment of repeated endoscopic biopsies, which is problematic. The use of biomarkers presents an opportunity to reduce sampling bias and improve our ability to risk-stratify these patients.

Optimal radiotherapy dose fractionation for spinal cord compression has not been defined. This retrospective study has compared patients treated in a single centre receiving either 1 or 2 fractions of radiotherapy or a fractionated course of more prolonged treatment for spinal cord compression. A total of 102 consecutive patients have been analysed, treated between January 1999 and June 2001. Sixty patients were male with a median age of 68 (range 32 to 90 years). The most common primary sites were breast (28%), prostate (28%) and lung (20%). The median duration of symptoms prior to diagnosis was 3 weeks (range 0 to 40); diagnosis within 24 hours of symptoms was achieved in only 13% and within one week in 29%. The diagnosis was confirmed on MRI in 93 patients (95%); the site of compression was dorsal spine (58%), lumbosacral (33%) and cervical cord (9%). Multiple levels of compression were present in 34 patients (33%). In 93 patients (91%) there were other sites of metastatic disease the most common of these being other bone metastasis. 51% were fully ambulant at diagnosis and 8% had complete paraplegia, the remainder having paraparesis with impaired mobility. At two months after RT the number of patients fully ambulant without significant neurological signs increased from 51% to 71% but the number of patients with complete paraplegia remained unchanged at 8%. A reduction in the incidence of impaired bladder function from 31% at presentation to 14% over the same time period was seen. Pain scores were derived for all patients on a simple retrospective 4 point scale. At presentation 59% recorded severe pain ( grade III) and 30% moderate pain (grade II). Only 4% had no pain (grade 0). At two weeks there was a substantial improvement in pain control with only 8% recording severe pain. Patients received a range of radiotherapy dose fractionation schedules the most common of which were 8Gy in a single fraction (22%) or 20Gy in 5 fractions (64%). In total 24 patients (23.5%) received a single or only 2 fractions (total dose 8 to 12Gy), the remainder receiving 20Gy in 5 fractions or in one case each 13Gy in 3 fractions, 13Gy in 4 fractions and 40Gy in 20 fractions. Patients receiving 1 or 2 fractions (1/2#) have been compared with the remainder (multi#). No difference in demographic parameters including primary tumour type and sites of cord compression was seen. A higher proportion of patients were paraplegic at presentation in the 1/2# (19%) than multi# (3%); when comparing outcome at two months from treatment the number of patients retaining mobility was statistically no different (80%,1/2# vs 68% multi#) and the proportion remaining paraplegic was unchanged (20%,1/2# vs 4% multi#). Pain control was also the same in both groups. In conclusion this series has confirmed that spinal cord compression should be diagnosed and treated urgently whilst the patient is mobile. Recovery after established paraplegia is rare. No advantage for protracted courses of radiotherapy over 1 or 2 fractions has been observed in this retrospective analysis. This data provides a strong basis for a prospective randomised trial evaluating dose fractionation in this condition. Alterations in DNA methylation probably play a major role in the development and progression of most, or all, tumour types. Many recent studies have identified increased methylation within the promoter regions of genes known to play important roles in cancer and demonstrated that such promoter hypermethylation was associated with loss of gene expression. Alteration in DNA methylation outside of promoter regions are also frequently observed, although the significance of such changes is less clear. We have previously demonstrated that aberrant promoter hypermethylation is common in ovarian cancer, targeting numerous genes, including many with known roles in tumour development (e.g. BRCA1) and resistance to chemotherapeutic agents (e.g. MLH1). Recently, the MCJ gene has been identified as a target for aberrant methylation in ovarian cancer and shown to play a role in sensitivity to several important anti-cancer drugs, such as cisplatin (Shridhar et al 2001, Cancer Res. 61:4258). Analysis of MCJ expression in our cell line models demonstrated that expression of the gene was lost in 8/10 cisplatin-resistant derivatives of the ovarian carcinoma cell line A2780. Furthermore, treatment of two of the resistant cell lines with 5-azacytidine, which inhibits DNA methyltransferase activity, resulted in re-expression of MCJ, suggesting that loss of expression may be due to increased methylation. Although the MCJ gene does not contain a classic CpG island within its promoter a previous report suggested that methylation of one particular CpG site within the promoter correlated with loss of expression. However, bisulfite sequencing analysis of this region, in the A2780 derivatives, identified no correlation between methylation of this, or other proximal CpG sites and gene expression. However, a CpG island was identified beginning within the first exon of MCJ, 164bp downstream of the transcriptional start site. Bisulfite sequencing of this region in DNA extracted from normal ovarian tissue determined that about 50% of clones were densely methylated and about 50% clones largely unmethylated. Sequencing of expressing cell lines revealed a pattern of methylation similar to normal DNA, whereas the cisplatinresistant, non-expressing, cell lines exhibited dense methylation of all clones. 5-azacytidine treatment of one of the non-expressing cell lines, which resulted in re-expression of MCJ, gave a methylation pattern similar to the MCJ expressing cell lines. These results suggest that methylation of the intragenic CpG island of MCJ may be responsible for loss of gene expression. Furthermore, sequencing of this region in ovarian tumour samples identified a subset of tumour (30%) which, similar to the nonexpressing ovarian cell lines, exhibited high levels of methylation of all clones. This raises the possibility that increased methylation of MCJ could play a role in controlling MCJ expression, and consequently in determining drug sensitivity, in ovarian cancer. A key step in the execution of apoptosis is the activation of caspases, one pathway of which involves the release of cytochrome c from mitochondria, which then binds to Apaf-1 leading to the activation of caspase-9. This, in turn, activates downstream execution caspases.
protein XIAP (X-linked Inhibitor of Apoptosis Protein). The suggested mechanism is an interaction between the N-terminal region of Smac and the BIR3 domain of XIAP, although there is also evidence that a Smac mutant that lacks the N-terminal amino acids still remains capable of potentiating apoptosis. Recently, it was shown that release of Smac from mitochondria is involved in dexamethasone-induced apoptosis in multiple myeloma cells and may a key link between the two pathways of caspase activation. We have constructed Ad CMV-Smac, a recombinant adenovirus encoding Smac/DIABLO. We demonstrate for the first time that delivery of the Smac gene to malignant cells can induce apoptosis: transfection of ovarian carcinoma cells with Ad CMV-Smac at multiplicities of infection of 3 -30 pfu/cell leads to increasing apoptosis in a dose-dependent manner. Western blot analysis confirms that Smac-induced apoptosis proceeds via a caspase-9 dependent pathway, which can be partially inhibited by the caspase-9 inhibitor zLEHD-fmk and by over-expression of XIAP. At later time points, however, there is also caspase-8 activation, which suggests that Smac may be able to activate multiple apoptotic pathways. We also show that Ad CMV-Smac can combine with other pro-apoptotic factors, such as cisplatin, paclitaxel and pro-caspase-3, to produce greater levels of apoptosis in transfected cells. Abnormalities in the cytochrome c/Apaf-1 pathway of caspase activation have been found in ovarian carcinoma cells. Furthermore, there is evidence that XIAP may be intimately linked to chemotherapy resistance in ovarian cancer and that down-regulation of XIAP can induce apoptosis in chemoresistant ovarian cancer cells. In light of this, we suggest that upregulation of Smac activity may have genuine potential in the treatment of ovarian cancer. Chronic myelogenous leukaemia (CML) is characterised at the molecular level by a (9;22) translocation which places the abl oncogene under the control of the bcr promoter generating a fusion protein (p210) with enhanced tyrosine kinase activity. CML cells are inherently resistant to apoptosis. The negative regulatory cytokine β-galactoside binding protein (β-GBP) induces cell cycle arrest at the late S/G2 threshold. Negation of this block via anti-βGBP antibodies allows normal cells to resume growth but neoplastic cells undergo apoptosis. Preliminary results indicate that lymphomas, mammary cell cancers and certain leukaemic cells undergo apoptosis following treatment with β-GBP at nanomolar concentrations, suggesting that βGBP may offer a novel and selective means of controlling neoplastic growth. We have examined the effect of βGBP on CML cell lines (K562, BV173, LAMA84, KYO-1) and have demonstrated apoptosis in 30-50% of the cells via TUNEL assays and Annexin V staining at 48 hours. Dual staining with propidium iodide, indicating cell cycle distribution, showed that at 24 hours cells exposed to βGBP had been arrested in S phase. Thus βGBP blocks the cell cycle in CML prior to activating cell death. These results prompted analysis of the effect of β-GBP on primary hemopoeitic progenitor cells (HPC's) from bone marrow donors and CML patients using in vitro colony forming assays. Treatment of HPC's from normal BM donors (n=22) with 400ng/ml βGBP did not significantly reduce CFU-GM number (median inhibition 0 -17%). By contrast treatment of HPC's from CML patients at diagnosis (n=13) resulted in a significant decrease in colony number (median inhibition 49-63%, p <0.005). Thus βGBP has a significant effect on the proliferative ability of leukemic progenitors while sparing normal hemopoeitic progenitor cells. Treatment of CML cells with βGBP also resulted in a significant decrease in the levels of the p210 BCR-ABL protein after 40-44 hours incubation and this downregulation preceded the appearance of significant numbers of apoptotic cells. Quantitation of bcr-abl mRNA levels by real time PCR using TaqMan technology indicated that mRNA levels are unchanged following treatment with βGBP indicating that P210 downregulation occurs at the post transcriptional level. p210 downregulation by βGBP is caspase-3 and proteasome independent. Thus two critical events, the prior introduction of a cell cycle block, allied to the subsequent downregulation of p210 BCR-ABL may both be required for apoptosis induction by βGBP in CML. The selective anti-proliferative and pro-apoptotic effect of βGBP on leukemic HPC's and its ability to downregulate p210 BCR-ABL may offer a new method to selectively induce apoptosis in CML cells while sparing normal hemopoietic progenitors. This may be useful in purging marrow in the context of autografting for CML or as an in vivo treatment, either alone or in combination with other chemotherapeutic agents. Thymidylate synthase (TS) is a critical chemotherapeutic target for fluoropyrimidines such as 5-fluorouracil (5-FU) and antifolates such as tomudex (TDX). The major limitations to TS-directed therapies are acquired or inherent resistance of tumour cells to these agents. Previous studies have demonstrated that TS expression levels are a key determinant of sensitivity of tumour cells to 5-FU and TDX. In order to identify more potential markers of response to TS-directed therapy, we have used DNA microarray technology to identify genes that are induced by 5-FU treatment in the MCF-7 breast cancer cell line. Of 2,400 genes analysed 30 were up-regulated by >8-fold. Of 10 highly up-regulated genes identified by the DNA microarray, only 6 were consistently found to be up-regulated in a dose and time dependent manner by Northern blot analysis. Consistently up-regulated genes included spermine/spermidine acetyl transferase (SSAT), annexin II, MAT8 and thymosin β-10. Analysis of other cell lines treated with 5-FU revealed that the induction of SSAT, annexin II, MAT8 and thymosin β-10 by 5-FU was not cell line specific. Furthermore, analysis of MCF-7 cells treated with TDX and oxaliplatin indicated that the expression of SSAT, annexin II, MAT8 and thymosin β-10 could be induced by both TS-specific and non-specific therapies. In addition, our results have shown that the expression of MAT-8 and thymosin β-10 were reduced significantly in a TDX resistant MCF-7 cell line compared to the parental line. Our results suggest that SSAT, annexin II MAT8 and thymosin β-10 may be involved in mediating the response of tumour cells to chemotherapy. Furthermore, our analysis demonstrates the potential of DNA microarrays to identify novel predictive markers and/or therapeutic targets.
Cyclo-oxygenase-2 (COX-2) expression is increased in breast cancer and surgery has been shown to increase the growth of metastatic tumours. The aim of our experiment was to investigate the effect of selective COX-2 inhibition on tumour metastases in an experimental excision model of breast cancer.50000 4T1 mammary carcinoma cells were injected into the mammary fat pad of 9 week old female BALB/c mice .On day 14 post injection (mean tumour size= 8.0+/-0.4 mm) the primary tumours were excised and the mice were randomised into 2 groups (n=7/group). One group received daily intra peritoneal (IP) injections of the selective COX-2 inhibitor, SC-236, the second group only received drug vehicle. Animals were sacrificed 14 days post excision of primary. At harvest there was a significant reduction in the number of lung metastases and serum VEGF levels between the SC-236 and control group. We also found a reduction in microvessel density and an increase in the apoptotic index. determined by the collection-rate method (2). Results show that the parent K562 and MCF cells exhibit significantly lower cytoplasmic conductivities of 0.026Sm -1 and 0.025Sm -1 , respectively than their drug resistant cells:-0.3 Sm -1 and 0.45Sm -1 , K562AR and MCF-7mdr, respectively. Furthermore, treatment of resistant cells with the Pgp specific MDR inhibitor XR 9576 showed a significant lowering of the internal conductivity to a value similar to that of the parental cells. A decrease in membrane permittivity was observed after treating K562AR with the modulator (from 10 to 6.9). In conclusion, We propose that this method offers great potential for the rapid assessment of the mechanisms of cancer drug action in vitro.as these data indicate that the electrical character of the cell is altered in MDR, but following modulation therapy, MDR cells achieve a similar profile to drug sensitive cells.
(1) Gascoyne, P.R.C. Wang, X. Huang, Y. Becker, F. 1997, IEEE Transactions on industry applications, Vol. 33 (3) An international multi-centre randomised controlled trial was conducted comparing whether the addition of neoadjuvant CMV to radical surgery or radiotherapy for patients with muscle-invasive transitional cell carcinoma of the bladder would improve survival. Eligible patients had histologically confirmed T2(G3), T3, T4a, N0/NX, M0 transitional cell carcinoma of the bladder and were judged suitable for curative treatment. CMV was given as 3 cycles consisting of 30mg/m 2 M & 4mg/m 2 V on days 1,8 and 100mg/m 2 C on day 2 of each cycle. From November 1989 to July 1995 a total of 976 patients were entered by 106 centres in 20 countries, 491 randomised to CMV and 485 to no CMV. 36% of patients were <60 years, 88% male, 69% WHO PS 0, 58% T3, 65% N0 and 88% G3. First results with a median follow-up of 4 years have been reported 1 in 1999 which showed a statistically non-significant 15% decrease in the risk of death after CMV (Hazard ratio 0.85 [95% CI 0.71-1.02] p=0.075). Currently a total of 559 patients have died and a final analysis is planned early 2002. At this time the median follow-up will be approximately 7 years. This is the largest trial making this comparison and the mature data will enable us to examine whether there is any evidence of CMV having a long-term effect on the main outcome of overall survival and other endpoints. Introduction: Cancer networks have the potential to make a powerful contribution to cancer research through collaborative studies. We conducted a feasibility study in the Northern and Yorkshire Region that attempted to recruit all eligible patients (pts.) into a randomised trial of adjuvant chemotherapy following radical local treatment for muscle invasive TCC of the bladder. Patients and Methods: The protocol was approved by Northern and Yorkshire Multicentre Research Ethics Committee. The target was to randomise 60 pts. within 2 years. Consenting pts. were registered at diagnosis. Following primary therapy (cystectectomy [CYST] or radical radiotherapy [RRT]), pts. were assessed for fitness to start 3 cycles of MVAC chemotherapy within 12 weeks. If suitable, they were given further information and randomised if they consented. Result: The trial was activated in 20 hospitals with enthusiastic support. 354 pts. were registered. 21% were not suitable for radical primary therapy due to metastatic disease or co-morbidity. Of the remainder, 50% had CYST and 50% had RRT. After CYST/RRT, 67% / 81% were medically unfit for chemotherapy. Of the 15% eligible for randomisation, 75% declined. The final number of patients randomised was 6. Conclusion: This study demonstrates that population-based cancer research within the NHS is feasible. In the patient group studied there was a higher incidence of co-morbidity than expected and many pts. declined randomisation. This experience is relevant to the development of the new National Cancer Research Network. The difficulty in recruitment in this area supports the recent decision by co-operative groups to collaborate in a international intergroup study to address the issue of adjuvant chemotherapy for bladder cancer. Using the Vysis GenoSensor Micro array System and the AmpliconI oncogene specific chip, we have screened 59 oncogenes for altered copy number in a cohort of 30 bladder patient samples consisting of 14 paired and 2 unpaired transitional cell carcinomas (TCC's). Following DNA extraction, nick translation and hybridisation, amplification profiles for each of the 59 arrayed oncogenes were generated. Using a stringent cut off value of 1.75, seven tumour samples showed normal copy number at all loci, while 9 tumours showed increased copy number of 1 (n=7) or multiple loci (n=2) involving 12 out of the 59 arrayed oncogenes when compared to paired normal samples. The genes identified were MDM2 (12q14.3), c-erbB-2 (17q), FGR (1p36), PIK3CA (3q26.3), ZNF127 (20q), AR (Xq) and JUNB (19p13.2). The c-MET (7q21) locus was the most commonly altered gene, with 3/16 (19%) tumours displaying increased copy number. The findings with c-MET are novel and significant for a number of reasons. Firstly, it functions as the receptor for the hepatocyte growth factor/ scatter factor (HGF/SF); a growth factor known to be a marker of disease stage and poor outcome in bladder cancer (Gohji, K., et al., 2000). Secondly, stimulation of c-met by its ligand HGF/SF leads to a whole range of biological downstream affects including scattering, angiogenesis, proliferation enhanced cell motility and invasion (Reviewed Maulik, G., et al., 2002). Finally, due to data already indicating the efficacy of receptor specific inhibitors to the c-kit and PDGFR receptor tyrosine kinases in gastrointestinal stromal tumours and NCLC tumours respectively, the possibility that c-MET may also be a suitable target for such antineoplastic therapy is clinically significant.   We present evidence that OBCAM (Opioid Binding Cell Adhesion Molecule) is a strong candidate tumor suppressor gene (TSG) inactivation of which is a fundamental event in the development of EOC. A frequent LOH rate of 51% was observed at the 11q25 marker D11S4085 in a series of 100 fresh and archive derived EOC normal/tumor pairs in an analysis using all polymorphic microsatellite markers from the region. In many cases, LOH demonstrated complete loss of an allele rather than allelic imbalance, suggesting lack of tumor heterogeneity. This indicated that LOH at D11S4085 is an early event in epithelial ovarian carcinogenesis. Bioinformatic analysis mapped D11S4085 within an intron of the OBCAM gene, a member of the IgLON family of GPI-anchored cell adhesion molecules. OBCAM may also be involved in signal transduction, and we have preliminary evidence that this is mediated through the raf/erk pathway. By quantitative RT-PCR, OBCAM expression is strong in normal human ovarian surface epithelium, but is extremely weak or undetectable in almost all primary ovarian tumours and cancer cell lines tested. Using Methylation Specific PCR, the OBCAM 5' CpG island is methylated in 81% (13/16) of ovarian cancer cell lines, and somatically methylated in 76% (32/42) of primary EOCs. In contrast, the OBCAM CpG island in normal ovary is unmethylated. In patients with LOH at D11S4085, i.e. within OBCAM, the second allele was somatically methylated in 81% of cases. Demethylation studies using 5'-aza 2'-deoxycytidine have demonstrated reexpression of OBCAM in a CpG island methylated, non-expressing clonal derivative of the ovarian cancer cell line, SKOV3.

MICROARRAY GENERATED ONCOGENE COPY NUMBER
OBCAM transfection into this methylated SKOV3 clonal derivative resulted in functional phenotypes in keeping with TSG function. OBCAM reexpression resulted in suppressed growth and enhanced aggregation of cells in vitro, markedly suppressed sub-cutaneous tumor growth in vivo and almost completely abolished intra-peritoneal tumorigenicity. OBCAM mutation analysis by SSCPE across approximately 200 ovarian primary tumors and cell lines has identified a single somatic mis-sense (Proline to Arginine) mutation within the first Ig domain in an ovarian tumor. As somatic mutation is an infrequent event, this supports abrogation of expression by LOH and by epigenetic mechanisms as the favoured means of OBCAM inactivation in epithelial ovarian cancer. The data presented is the first description of involvement of the IgLON family in cancer.

REAL TIME QUANTITATIVE POLYMERASE CHAIN REACTION ANALYSIS OF HPV 16 IN ADENOCARCINOMA OF CERVIX
GK Chew 1 *, ME Cruickshank 1 , PH Rooney 2 , DE Parkin 1 and GI Murray 2 . 1 Department of Gynaecology-Oncology, Aberdeen Royal Infirmary, Aberdeen. 2 Department of Pathology, University of Aberdeen, Aberdeen.
There has been a relative and absolute increase in the incidence of adenocarcinoma of the cervix (ACC), particularly in the younger women. ACC represents 15-20% of all cervical cancer and will become more prominent as the incidence of squamous cell carcinoma of the cervix (SCC) continues to fall with organised cervical screening. Both HPV 16 and 18 are associated with ACC. In some in-situ polymerase chain reaction (PCR), HPV 16 was noted to be present in the adjacent uninvolved cervical epithelium. In this study, we have coupled two powerful techniques, laser capture microdissection and real-time quantitative PCR to determine the HPV16 prevalence in the women with ACC and to accurately assess the HPV16 DNA copy number in the cells. The association between HPV16 positivity and the age of the women, grade and stage of the cancer was analysed.A cohort of women diagnosed with ACC between 1991-2000 inclusive was identified. The histology was reviewed to confirm the diagnosis and adenosquamous carcinomas were excluded. The ACC cells were isolated from archival paraffin-fixed 5µm sections using a PixCell II laser microdissection system (Arcturus Engineering). Tumour tissue was identified and removed by the laser onto a transfer film. Approximately 200 laser pulses were taken per tumour specimen. DNA extraction was carried out by incubating the dissected cells in a 3mg/ml proteinase K at 55°C for 4 hours and then heat inactivated. DNA was assessed by Taqman PCR using the ABI7700 (PE Applied Biosystems). Real time quantitative PCR was used to amplify the internal control gene, Beta-globin and the HPV 16 gene. The Caski cell line is known to have 600 copies HPV16 DNA per cell. DNA was extracted from a known number of Caski cells and serially diluted to 10 -9 concentration. Real-time quantitative PCR was used to determine the Ct count for Beta-globin and HPV 16 at the different concentrations in order to generate a standard curve. For each case, the HPV16 and Beta-globin gene copy number was calculated using the mean values of Ct (each case was assayed in triplicate) and plotted against the standard curve to determine the gene copy number.55 women were identified to have ACC cervix from 1991-2000 inclusive. The age distribution of this cohort showed a bimodal pattern, with peaks in the 36-45 and >55 years age groups. In 3 cases, the DNA extraction was unsuccessful and these cases were excluded. Of the remaining 52 women, 25% were HPV 16 positive. HPV positivity was more common in the well and moderately differentiated tumours and in women less than 45 years old. There is no difference in the HPV copy number in the grade and stage of the cancer. The HPV 16 gene copy number is significantly higher in women more than 45 years old (p=0.03). Between March 1996 andMarch 2001, 96 patients (pts.), aged 27-90 (median 53 years) with recurrent or refractory follicular lymphoma (FL) 64 pts., mantle cell lymphoma 5 pts., lymphoplasmacytic lymphoma (LPC) 4 pts., small lymphocytic lymphoma 3 pts. and a further 20 pts. who had undergone transformation (Tx) to large B cell lymphoma (LBCL, 19 pts. previous FL, 1 pt. LPC) were enrolled to be treated either in an open phase II trial (61 pts.) or on a compassionate basis (35 pts.) with Bexxar™(tositumomab and I 131 tositumomab). Forty patients had bone marrow infiltration (all <25%). The median number of previous therapies was 2 (range 1-9). Eleven pts. had progressed following previous high-dose therapy (HDT), Rituximab (RXB) 13pts., or Bexxar™ 3 pts. Administration, after dosimetry, was as previously described, delivering a whole body dose of 75cGy if the platelet (plt.) count was >150x10 9 /l, 65 cGy for pts. with a plt. count of 100-149x10 9 /l and 45cGy following HDT. Seven pts. did not receive therapy; 3 because of human antimouse antibodies, 1 pt. withdrew and 3 progressed after dosimetry. Response was first evaluated 7 weeks post therapy and repeated 3 monthly. Further disease regressions were observed for upto 1 year, although progressions occurred in other pts. during this time. Toxicity was principally haematological. Median neutrophil and plt. nadirs occured at 7 and 6 weeks respectively. Grade 3 or 4 toxicity: neutropenia in 38% of pts. and thrombocytopenia in 31%. With a median follow up of 2 1 /4 years, the median duration of response was 2 1 /4 years (95% CI:1.1 to not reached) for those in CR/CR(u) and 7 months (3 mo. to not reached) for those in PR. Overall response rate at first evaluation declined with successive number of previous therapies: 73%, 66% and 46% respectively for 1, 2 or 3 or more treatments. The efficacy of Bexxar™ has been demonstrated in a range of clinical settings with the longest remission duration in those who achieve CR/CR(u). Supported by the ICRF, CRC, Corixa Corp and the NHS. PURPOSE: To assess the safety, tolerability, efficacy, immune stimulatory and anti-angiogenic effects of a novel thalidomide analogue, CDC-501 (REVIMIDä), in the treatment of patients with advanced cancer. PATIENTS AND METHODS: Twenty patients with heavily pre-treated advanced stage IV malignant melanoma (n=13), pancreatic (n=2) and other cancers (n=5) were treated with a dose escalating regimen of CDC-501 starting at 5mg and rising to 50mg after 4 weeks. Clinical efficacy, tolerance and adverse effects were evaluated. In vitro analysis of peripheral T cell surface markers and serum for cytokines and pro-angiogenic factors were performed. RESULTS: CDC-501 was generally well tolerated and no serious adverse events were attributed to its use. Fourteen patients completed the study, 3 withdrew due to disease progression and 3 due to withdrawal of consent. CDC-501 treatment led to clinical responses in a number of patients. All patients showed evidence of activation of both CD4+ and CD8+ T cells as measured by increased CD45RO+ expression. Furthermore, activation of memory cells (CD45RA+/L-selectinlow) in some patients may indicate activation of tumourspecific cells. This was associated with increased serum levels of GM-CSF, sIL-2Rα, IL-12 and TNF-α. However, levels of pro-angiogenic factors were unchanged. CONCLUSION: This study demonstrates the safety and significant clinical efficacy of CDC-501 in the treatment of recalcitrant advanced cancer. The induction of immune activation suggests a more beneficial role for this class of compound in less advanced patients as an immunostimulatory adjuvant perhaps combined with tumour cell vaccination regimens. The most common primary brain tumours are glial tumours, including the highly malignant glioblastoma. With current treatments of surgery, radiotherapy and chemotherapy outcome remains poor with a median survival of one year. A novel strategy to combat this disease is the utilisation of the oncolytic herpes simplex virus, HSV1716. HSV1716 selectively replicates in actively dividing cells. Two phase I clinical trials using HSV1716 have been completed. In the first, patients with recurrent glioma received direct intratumoural injection of escalating virus dose. In the second, patients with recurrent as well as newly diagnosed high grade glioma were injected with HSV1716 into the tumour prior to resection at 5-8 days. No toxicity was seen in any patients and