To date,nearly 100 patients with thoracic malignancies (mostly lung cancers) have received deoxyazacytidine (DAC) Depsipeptide (DP), or sequential DAC/DP infusions on protocols initiated in the Thoracic Oncology Section. Clinical toxicities, and response to therapy have been assessed by CTCAE and RECIST criteria, respectively. Plasma DAC and DP levels have been evaluated by LC-MS and HPLC techniques. Quantitative RT-PCR, methylation-specific-PCR, immunohistochemistry, and ELISA techniques have been used to assess a variety of molecular endpoints in pre-and post-treatment tumor biopsies and sera. Micro-array techniques have been used to comprehensively examine gene expression profiles in laser-captured tumor cells from pre- and post treatment biopsies from 21 individuals receiving DAC, DP, or sequential DAC/DP infusions. Results of these arrays have been compared to data derived from analysis of laser-captured tumor cells and adjacent, histologically normal bronchial epithelia from 20 patients undergoing definitive lung cancer resections. Whereas no objective clinical responses have been observed, several patients have exhibited prolonged stabilization of disease following DAC, DP, or sequential DAC/DP infusions. Plasma DAC and DP concentrations have approximated threshold levels for gene induction and apoptosis in cultured lung cancer cells. Approximately 30% of patients receiving DAC or DP infusions have exhibited enhanced expression of NY-ESO-1, p16, p21, or acetylated histones H3 and H4 in tumor biopsies. Long-oligo array analyses have revealed complex, heterogeneous responses to DAC, DP, and DAC/DP in lung cancer cells, with an apparent shift of gene expression profiles toward those observed in histologically normal bronchial epithelia. Whereas coordinate activation of BORIS variant, stem cell, and CT-X genes in cancer cells suggests either direct or functional interactions of their respective gene products during germ cell development, the frequency and clinical relevance of activation of these genes during pulmonary carcinogenesis have not been examined in a comprehensive manner. Recently, experiments have been initiated to comprehensively examine BORIS variant- relative to stem cell, and CT-X gene expression profiles using RNA and DNA from lung cancer specimens in the Ludwig Cancer Institute/CUMC lung cancer bank. Clinical-pathologic data including tumor histology and stage, treatment regimens, and outcome for the respective patients, all of whom have undergone surgery at New York-Presbyterian Hospital, have been entered into a database at CUMC. Because the DNA and RNA have been derived from whole tumors, not laser captured cancer cells, techniques will be established to assess mRNA copy numbers for BORIS variant, stem cell, and CT-X genes relative to thyroid transcription factor 1 (TTF1), which is expressed in the majority of pulmonary adenocarcinomas, but not in normal lung, or inflammatory cells. Data from these array experiments will be merged with the CUMC clinical database. Results of subsequent analysis should establish the frequency of, relationships between, and prognostic significance of BORIS variant, stem cell, and CT-X gene activation in pulmonary adenocarcinomas. A 2008 NIH Bench-to-Bedside Award (PI- DSS) will directly support these experiments. One potential strategy to overcome limitations of immunotherapy for pulmonary malignancies related to low level, heterogeneous intratumoral CT-X antigen expression, and the lack of recombinant polyvalent vaccines, involves vaccination of patients with cancer lines exhibiting high-level expression of multiple potential targets. We have identified several established cancer lines such as H1299 which exhibit high-level expression of BORIS variants and CT-X genes that could be considered for allogeneic vaccine development. An alternative strategy involves the use of epigenetically-modified autologous tumor cells to immunize lung cancer patients against a variety of CT-X antigens that potentially can be up-regulated in their respective primary cancers by systemic gene induction regimens; to date, no such efforts have been reported. As such, a series of experiments were conducted to examine the potential feasibility of using epigenetically-modified autologous tumor cells for cancer vaccines. Briefly, tissues/fluids from patients with tumors of various histologies were processed for primary culture. Sources of tumor cells included peritoneal fluid, endoscopic pleural or mediastinal biopsies, as well as CT-guided FNAs. Several of the recently derived cancer lines were treated with sequential DAC/DP under exposure conditions approximating or greatly exceeding those achievable in our current gene induction trials. These experiments revealed heterogeneous robust CT-X gene induction, suggesting that in vitro manipulation of CT-X gene expression in short-term cancer lines is feasible. A 2008 NIH Bench-to-Bedside Award will directly support an autologous cancer cell vaccine trial, as well as generation, and FDA certification of a master bank of H1299 lung cancer cells which can be used as an allogeneic cell vaccine in novel protocols initiated in the Thoracic Oncology Section. Micro-array analysis of biopsies from patients on our phase II DP (Romidepsin) trial suggested that this HDAC inhibitor diminishes Aurora kinase expression in lung cancer cells. As such, a series of experiments were conducted to examine this issue.Preliminary quantitative RT-PCR experiments demonstrated that Aurora A and B mRNA levels in lung cancer cells were considerably higher than levels in normal pulmonary epithelia. Additional quantitative RT-PCR and promoter-reporter assays revealed that DP, as well as TSA and SAHA inhibited Aurora A, Aurora B, and survivin expression with kinetics that were remarkably similar within individual cell lines, and appeared to coincide with p53 expression status. These effects were not consistently observed following treatment with geldanamycins. Western blot experiments indicated that the kinetics of Aurora kinase and survivin depletion were also contingent on the relative stabilities of these proteins in lung cancer cells following DP exposure. ChIP experiments indicated that inhibition of Aurora B promoter activity coincided with decreased acetylated H3K9 and dimethyl H3K4 (activation marks), and a concommitant increase in trimethylated H3K9 (repression mark), as well as recruitment of MBD1, MBD2, and MBD3 to the minimal Aurora B promoter. Interestingly, recruitment of MBD3 to the ErbB2 promoter coincides with TSA-mediated down-regulation of this oncogene in lung cancer cells. Confocal imaging experiments demonstrated that DP and TSA decreased expression and altered localization of Aurora A, Aurora B, and survivin, resulting in mitotic catastrophe in lung cancer cells. Collectively, these data indicate that novel transcriptional regulatory mechanisms mediating Aurora kinase and survivin expression may contribute to cytotoxicity mediated by HDAC inhibitors in lung cancer cells.
Showing the most recent 10 out of 26 publications
