Gemcitabine plus sorafenib versus gemcitabine alone in advanced biliary tract cancer: a double-blind placebo-controlled multicentre phase II AIO study with biomarker and serum programme

BACKGROUND

Since sorafenib has shown activity in different tumour types and gemcitabine regimens improved the outcome for biliary tract cancer (BTC) patients, we evaluated first-line gemcitabine plus sorafenib in a double-blind phase II study.

PATIENTS AND METHODS

102 unresectable or metastatic BTC patients with histologically proven adenocarcinoma of gallbladder or intrahepatic bile ducts, Eastern Cooperative Oncology Group (ECOG) 0-2 were randomised to gemcitabine (1000 mg/m2 once weekly, first 7-weeks+1-week rest followed by once 3-weeks+1-week rest) plus sorafenib (400 mg twice daily) or placebo. Treatment continued until progression or unacceptable toxicity. Tumour samples were prospectively stained for sorafenib targets and potential biomarkers. Serum samples (first two cycles) were measured for vascular endothelial growth factors (VEGFs), vascular endothelial growth factor receptor 2 (VEGFR-2) and stromal cell-derived factor 1 (SDF1)α by enzyme-linked immunosorbent assay (ELISA).

RESULTS

Gemcitabine plus sorafenib was generally well tolerated. Four and three patients achieved partial responses in the sorafenib and placebo groups, respectively. There was no difference in the primary end-point, median progression-free survival (PFS) for gemcitabine plus sorafenib versus gemcitabine plus placebo (3.0 versus 4.9 months, P=0.859), and no difference for median overall survival (OS) (8.4 versus 11.2 months, P=0.775). Patients with liver metastasis after resection of primary BTC survived longer with sorafenib (P=0.019) compared to placebo. Patients who developed hand-foot syndrome (HFS) showed longer PFS and OS than patients without HFS. Two sorafenib targets, VEGFR-2 and c-kit, were not expressed in BTC samples. VEGFR-3 and Hif1α were associated with lymph node metastases and T stage. Absence of PDGFRβ expression correlated with longer PFS.

CONCLUSION

The addition of sorafenib to gemcitabine did not demonstrate improved efficacy in advanced BTC patients. Biomarker subgroup analysis suggested that some patients might benefit from combined treatment.

Comprehensive allelotype and genetic anaysis of 466 human nervous system tumors

Brain tumors pose a particular challenge to molecular oncology. Many different tumor entities develop in the nervous system and some of them appear to follow distinct pathogenic routes. Molecular genetic alterations have increasingly been reported in nervous system neoplasms. However, a considerable number of affected genes remain to be identified. We present here a comprehensive allelotype analysis of 466 nervous system tumors based on loss of heterozygosity (LOH) studies with 129 microsatellite markers that span the genome. Specific alterations of the EGFR, CDK4, CDKN2A, TP53, DMBT1, NF2, and PTEN genes were analyzed in addition. Our data point to several novel genetic loci associated with brain tumor development, demonstrate relationships between molecular changes and histopathological features, and further expand the concept of molecular tumor variants in neuro-oncology. This catalogue may provide a valuable framework for future studies to delineate molecular pathways in many types of human central nervous system tumors.

Differences in allelic distribution of two polymorphisms in the VHL-associated gene CUL2 in pheochromocytoma patients without somatic CUL2 mutations

Although the two major familial forms of pheochromocytomas, multiple endocrine neoplasia type 2 and von-Hippel-Lindau disease (VHL), have been associated with mutations of the RET and VHL genes, respectively, the molecular pathogenesis of sporadic pheochromocytomas is largely unknown. Recently, a putative tumor suppressor gene has been identified, CUL2, whose product has been shown to interact with the VHL tumor suppressor. To examine whether CUL2 plays a role in pheochromocytoma pathogenesis, we analyzed a series of 26 distinct tumor samples for mutations in the whole coding region of this gene. There were no somatic pathogenic mutations in CUL2, except for 1 sporadic tumor that had a hemizygous gene deletion. We also found 3 novel polymorphisms in the gene. One of these variants, IVS5-6C/T, as well as another previously described one, c.2057G/A, were overrepresented among the pheochromocytoma patients compared to that in a control population (P < 0.005 and P < 0.01, respectively). Although our findings suggest that CUL2 does not play a major role in the pathogenesis of pheochromocytomas, it is still unknown whether epigenetic mechanisms are involved in its inactivation in VHL-associated tumors. Furthermore, the potential role for the overrepresented alleles in the pheochromocytoma group requires further investigation.

PTEN mutations in gliomas and glioneuronal tumors

Cytogenetic and loss of heterozygosity studies have suggested the presence of at least one tumor suppressor gene on chromosome 10 involved in the formation of high grade gliomas. Recently, the PTEN gene, also termed MMAC1 or TEP1, on chromosomal band 10q23 has been identified. Initial studies revealed mutations of PTEN in limited series of glioma cell lines and glioblastomas. In order to systematically evaluate the involvement of PTEN in gliomas, we have analysed the entire PTEN coding sequence by SSCP and direct sequencing in a series of 331 gliomas and glioneuronal tumors. PTEN mutations were detected in 20/142 glioblastomas, 1/7 giant cell glioblastomas, 1/2 gliosarcomas, 1/30 pilocytic astrocytomas and 2/22 oligodendrogliomas. No PTEN mutations were detected in 52 astrocytomas, 37 oligoastrocytomas, three subependymal giant cell astrocytomas, four pleomorphic xanthoastrocytomas, 15 ependymomas, 16 gangliogliomas and one dysembryoplastic neuroepithelial tumor. In addition, all tumors were examined for the presence of homozygous deletions of the PTEN gene; these were detected in 7 glioblastomas that did not have PTEN mutations. Therefore, PTEN mutations occur in approximately 20% of glioblastomas but are rare in lower grade gliomas. These findings confirm that PTEN is one of the chromosome 10 tumor suppressor genes involved in the development of glioblastomas.

Analysis of the PTEN gene in human meningiomas

Previous observations demonstrated that the neurofibromatosis type 2 gene (NF2) plays an important role in the pathogenesis of the transitional, fibroblastic and malignant variants of human meningiomas. No specific genes have been associated with the pathogenesis of meningothelial meningiomas and with the progression to anaplastic meningiomas. However, allelic losses on chromosomal arms 1p, 10q and 14q have been implicated in the process of malignant progression. Recently, PTEN (phosphatase and tensin homolog deleted on chromosome ten) also termed MMAC1 (mutated in multiple advanced cancers 1) or TEP1 (TGF--regulated and epithelial cell-enriched phosphatase), emerged as a candidate gene on chromosome 10q23.3. Initial studies revealed mutations of PTEN in limited series of glioblastomas, breast, kidney and prostate carcinomas mainly as cell lines. In order to evaluate the involvement of PTEN in the development of meningiomas, we have analysed the entire coding sequence of the gene in a series of 55 meningiomas (WHO grade I). 10 atypical meningiomas (WHO grade II) and 10 anaplastic meningiomas (WHO grade III). No PTEN mutations were seen in the WHO grade I meningiomas. However, one of the anaplastic meningiomas carried a somatic mutation. In addition, all tumours were examined for the presence of homozygous deletions of PTEN but these were not detected in any of the meningiomas. Our data suggest that mutations in PTEN are not involved in the formation of low grade meningiomas, but may contribute to malignant progression in a fraction of anaplastic meningiomas.