Johannes Merk

Establishment of Patient-Derived Non–Small Cell Lung Cancer Xenografts as Models for the Identification of Predictive Biomarkers

Purpose: It was the aim of our study to establish an extensive panel of non-small cell lung cancer (NSCLC) xenograft models useful for the testing of novel compounds and for the identification of biomarkers. Experimental Design: Starting from 102 surgical NSCLC specimens, which were obtained from primarily diagnosed patients with early-stage tumors (T2/T3), 25 transplantable xenografts were established and used for further investigations. Results: Early passages of the NSCLC xenografts revealed a high degree of similarity with the original clinical tumor sample with regard to histology, immunohistochemistry, as well as mutation status. The chemotherapeutic responsiveness of the xenografts resembled the clinical situation in NSCLC with tumor shrinkage obtained with paclitaxel (4 of 25), gemcitabine (3 of 25), and carboplatin (3 of 25) and lower effectiveness of etoposide (1 of 25) and vinorelbine (0 of 11). Twelve of 25 NSCLC xenografts were >50% growth inhibited by the anti-epidermal growth factor receptor (EGFR) antibody cetuximab and 6 of 25 by the EGFR tyrosine kinase inhibitor erlotinib. The response to the anti-EGFR therapies did not correlate with mutations in the EGFR or p53, but there was a correlation of K-ras mutations and erlotinib resistance. Protein analysis revealed a heterogeneous pattern of expression. After treatment with cetuximab, we observed a down-regulation of EGFR in 2 of 6 sensitive xenograft models investigated but never in resistant models. Conclusion: An extensive panel of patient-derived NSCLC xenografts has been established. It provides appropriate models for testing marketed as well as novel drug candidates. Additional expression studies allow the identification of stratification biomarkers for targeted therapies.

Correlation of SHOX2 Gene Amplification and DNA Methylation in Lung Cancer Tumors

BackgroundDNA methylation in the SHOX2 locus was previously used to reliably detect lung cancer in a group of critical controls, including cytologically negative samples with no visible tumor cell content, at a high specificity based on the analysis of bronchial lavage samples. This study aimed to investigate, if the methylation correlates with SHOX2 gene expression and/or copy number alterations. An amplification of the SHOX2 gene locus together with the observed tumor-specific hypermethylation might explain the good performance of this marker in bronchial lavage samples.MethodsSHOX2 expression, gene copy number and DNA methylation were determined in lung tumor tissues and matched morphologically normal adjacent tissues (NAT) from 55 lung cancer patients. Quantitative HeavyMethyl (HM) real-time PCR was used to detect SHOX2 DNA methylation levels. SHOX2 expression was assayed with quantitative real-time PCR, and copy numbers alterations were measured with conventional real-time PCR and array CGH.ResultsA hypermethylation of the SHOX2 locus in tumor tissue as compared to the matched NAT from the same patient was detected in 96% of tumors from a group of 55 lung cancer patients. This correlated highly significantly with the frequent occurrence of copy number amplification (p < 0.0001), while the expression of the SHOX2 gene showed no difference.ConclusionsFrequent gene amplification correlated with hypermethylation of the SHOX2 gene locus. This concerted effect qualifies SHOX2 DNA methylation as a biomarker for lung cancer diagnosis, especially when sensitive detection is needed, i.e. in bronchial lavage or blood samples.

Patient-derived xenografts of non-small-cell lung cancer: a pre-clinical model to evaluate adjuvant chemotherapy? §

OBJECTIVEnRecent trials have evaluated adjuvant chemotherapy in patients with non-small-cell lung cancer (NSCLC). For stage IB to IIIA, a significant improvement of treatment results for platin-based chemotherapy was shown, but only one of the 20 patients treated has a benefit of disease-free and overall 5-year survival. In future the implementation of biomarkers, novel agents and individual selection may contribute to better treatment results in adjuvant therapy. Pre-clinical models are one way to study treatment innovations.nnnMATERIALS AND METHODSnWe have developed a lung cancer xenograft model. Fresh tumour material of patients with NSCLC was subcutaneously transplanted to immunodeficient mice shortly after surgical resection. In total, 102 samples have been transplanted from which 25 passagable models could be generated. Of the established xenograft lines, 48% were derived from squamous cell carcinomas and 24% from adenocarcinomas. All but one originated from long-term smokers.nnnRESULTSnIt could be shown that the early murine passages (maximum 10) were similar to the original tumour with regard to histology and the expression of the surface proteins as E-cadherin, EpCAM or the cell proliferation marker Ki-67. The growth rate of the established xenografts was a unique feature of the different models and not related to patient characteristics or to the histology type. All xenograft models showed a wide variability in response to both classical chemotherapy and targeted anti-epidermal growth factor receptor agents. Response rates were in good accordance with the results of recent clinical studies.nnnDISCUSSIONnIn summary, we have developed a panel of patient-derived NSCLC xenografts. These xenograft models could be used for pre-clinical studies to evaluate chemotherapy, novel targeted therapies and expression of potential biomarkers.

Comparative Profiling of the Novel Epothilone, Sagopilone, in Xenografts Derived from Primary Non–Small Cell Lung Cancer

Purpose: Characterization of new anticancer drugs in a few xenograft models derived from established human cancer cell lines frequently results in the discrepancy between preclinical and clinical results. To take the heterogeneity of tumors into consideration more thoroughly, we describe here a preclinical approach that may allow a more rational clinical development of new anticancer drugs. Experimental Design: We tested Sagopilone, an optimized fully synthetic epothilone, in 22 well-characterized patient-derived non–small cell lung cancer models and correlated results with mutational and genome-wide gene expression analysis. Results: Response analysis according to clinical trial criteria revealed that Sagopilone induced overall responses in 64% of the xenograft models (14 of 22), with 3 models showing stable disease and 11 models showing partial response. A comparison with response rates for established drugs showed the strong efficacy of Sagopilone in non–small cell lung cancer. In gene expression analyses, Sagopilone induced tubulin isoforms in all tumor samples, but genes related to mitotic arrest only in responder models. Moreover, tumors with high expression of genes involved in cell adhesion/angiogenesis as well as of wild-type TP53 were more likely to be resistant to Sagopilone therapy. As suggested by these findings, Sagopilone was combined with Bevacizumab and Sorafenib, drugs targeting vascular endothelial growth factor signaling, in Sagopilone-resistant models and, indeed, antitumor activity could be restored. Conclusion: Analyses provided here show how preclinical studies can provide hypotheses for the identification of patients who more likely will benefit from new drugs as well as a rationale for combination therapies to be tested in clinical trials. Clin Cancer Res; 16(5); 1452–65

Chemoresistance in non-small-cell lung cancer: can multidrug resistance markers predict the response of xenograft lung cancer models to chemotherapy?

OBJECTIVEnIn chemotherapy for non-small-cell lung cancer (NSCLC), some patients seem to exhibit an intrinsic resistance or develop an acquired resistance under treatment. Results on resistance markers for possible treatment failure as shown in studies on selected lung cancer cell lines could not be completely confirmed in clinical trials. As these conflicting data require further research, we created a model between cell culture and the clinical need to study this problem.nnnMETHODSnOur study was based on patient-derived NSCLC xenografts in a mouse model, which revealed a high coincidence with the original tumour. Protein and messenger RNA (mRNA) expression of known resistance markers (breast cancer resistance protein (BCRP), multidrug resistance P-glycoprotein (MDR), lung cancer-related protein (LRP) and multidrug resistance protein 1 (MRP1)) were analysed by real-time polymerase chain reaction (PCR) and immunoblotting in 24 xenografts. Chemosensitivity to etoposide, carboplatin, gemcitabine, paclitaxel, cetuximab and erlotinib was determined in in vivo xenograft experiments and compared with the protein and mRNA expression of the multidrug resistance markers.nnnRESULTSnWith the exception of a single correlation between chemosensitivity and mRNA expression of etoposide and bcrp (mRNA expression of BCRP), we found no significant correlation between the response rates and protein- and mRNA expression levels in our 24 xenografts. The present results indicate that in vivo expression levels of multidrug resistance proteins and their mRNAs may not play a comparable role in chemoresistance of NSCLC, as pointed out in selected tumour cell lines.nnnCONCLUSIONSnPatient-derived xenografts allow detailed investigation of therapy-related markers and their dynamic regulation in a well-standardised and clinically related way. As a consequence of our investigations, we regard multidrug resistance to be a multifactorial phenomenon, in which more factors than the markers analysed by the present study may be involved.

Response of Patient-Derived Non-Small Cell Lung Cancer Xenografts to Classical and Targeted Therapies Is Not Related to Multidrug Resistance Markers

Tumor cells that are nonsensitive to anticancer drugs frequently have a multidrug resistant (MDR) phenotype. Many studies with cell lines and patient material have been done to investigate the impact of different resistance markers at protein and mRNA level in drug resistance but with contradictory outcome. In the present study, 26 well-characterised patient-derived non-small cell lung cancer xenografts were used. The known chemosensitivity to etoposide, carboplatin, gemcitabine, paclitaxel and erlotinib was compared to the protein and mRNA expression of BCRP, LRP, MDR1, and MRP1. Further, four of these xenografts were short-term treated to analyse possible regulation mechanisms after therapeutic interventions. We found a borderline correlation between the bcrp mRNA expression and the response of xenografts to etoposide. All other constitutive mRNA and protein expression levels were not correlated to any drug response and were not significantly influenced by a short term treatment. The present results indicate that the expression levels of MDR proteins and mRNA investigated do not play an important role in the chemoresistance of NSCLC in the in vivo situation.

Activation of AMP-activated protein kinase sensitizes lung cancer cells and H1299 xenografts to erlotinib

OBJECTIVESnThe therapeutic scheme for non-small cell lung cancer (NSCLC) patients can be improved if adapted to the individual response. For example, 60-70% of adenocarcinoma patients show response to EGFR-tyrosine kinase inhibitors in the presence of mutated EGFR. We searched for additional target molecules involved in the action of the EGFR-tyrosine kinase inhibitor erlotinib in the absence of EGFR mutations, which might be suitable for combinatorial therapy approaches.nnnMATERIALS AND METHODSnErlotinib-response associated proteins were investigated in patient-derived NSCLC mouse xenografts by reverse-phase protein array technology (RPPA) and Western blotting. A combinatorial treatment approach was carried out in NSCLC cell lines and H1299 mouse xenografts, and subsequently analyzed for consequences in cell growth and signal transduction.nnnRESULTSnAMP-activated protein kinase (AMPK) expression was increased in erlotinib responders before and after treatment. In a combinatorial approach, activation of AMPK by A-769662 and erlotinib treatment showed a synergistic effect in cell growth reduction and apoptosis activation in H1299 cells compared to the single drugs. AMPK pathway analyses revealed an effective inhibition of mTOR signaling by drug combination. In H1299 xenografts, the tumor size was significantly decreased after combinatorial treatment.nnnCONCLUSIONnOur results suggest that AMPK activation status affects response to erlotinib in distinct lung tumor models.

Abstract 4524: Characterization of 28 patient derived colorectal cancer xenografts and analysis for response markers to an antiangiogenic treatment

Proceedings: AACR 103rd Annual Meeting 2012‐‐ Mar 31‐Apr 4, 2012; Chicago, ILnnWe have recently described an extended panel of new primary human CRC xenografts and their preclinical chemo-sensitivity to anti-cancer drugs (Poster 1584/3 AACR Annual Meeting 2010). These models consider more thoroughly the heterogeneity and genetic instability of tumors in a way ultimately enabling treatment decisions on an individualized basis. The approach is based on large scale screening for activity in new in vivo models combined with extensive characterization of these models (gene expression and mutations). Bioinformatic methods are used to further evaluate the mechanism of action of new drugs in colorectal cancer, to investigate possible resistance mechanisms and provide a rationale for combination therapies. Here we describe the evaluation of an antiangiogenic treatment in this panel of primary patient derived colorectal cancer models. Primary colon carcinoma tissue samples were collected using a standardized procedure. Tumor pieces were transplanted into immunodeficient mice immediately after surgery. A panel of 28 stably passagable colon cancer xenografts could be established as permanent tumor models. The results obtained so far, show that these patient-derived colon cancer models feature a high coincidence with the original tumor regarding histology and gene expression profiling. Response analysis in a subset of 19 xenografts evaluating the treated-to-control (T/C) ratios of tumor volumes revealed that the human specific VEGF antibody bevacizumab induces a meaningful biological effect (T/C below 50%) in 13 of 19 models. A first comparative genome-wide gene expression analyses of this subset revealed reasonable genes and pathways possibly involved in bevacizumab sensitivity. We will extend this investigation to all 28 tumor models in order to enhance the statistical robustness of the preliminary data. Our approach demonstrates that analysis of drug sensitivity together with bioinformatics correlations can provide candidate markers for further studies. An early implementation of such an approach in the drug development process could be helpful for the identification of predictive markers helping to select tumors likely to benefit from the drug as well as provide a rationale for combinations to be tested in clinical trials. These new strategies may help to better integrate research results in clinical development and may increase the success of new anti-cancer drugs in clinical studies.nnCitation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4524. doi:1538-7445.AM2012-4524

Abstract A14: Patient-derived tumor xenografts supporting development of targeted drugs.

Almost in parallel with the new millennium processes in drug development for cancer therapy have changed substantially. Insight in the cancer biology have revealed this disease as a complex genetic disorder leading to a high variety of phenotypes among the different individuals. Many tumors present a very specific pattern of molecular changes and this leads to a heterogenous individual response to drugs. The correct identification of predictive biomarkers selecting the most appropriate therapies and avoiding unnecessary treatments for an individual patient is still a challenge. Patient-derived tumor xenografts allow preclinical investigations in a clinically relevant way. We performed investigations to improve the understanding of cancer complexity and to draw rational conclusions for therapy decisions. Tumor models were established by direct transplantation of surgical specimens to immunodeficient mice and were maintained in early passages. We have shown a high correlation between original patient sample and xenograft both at gene and protein level. We established several large panels of tumor models: i.e. 10 breast, 30 colorectal, 25 lung, 6 ovarian, 10 sarcomas, and 30 leukemias (25 ALL and 5 AML). During the characterization of the xenografts for response towards clinically used anticancer drugs, the analysis for mutations and of gene expression revealed interesting correlations. In NSCLC models upregulation of hypoxia marker genes correlated with resistance to microtubule targeted drugs. In the colon cancer xenografts KRAS mutations predicted resistance to cetuximab. We further identified a potential set of 20 genes which may be predictive for oxaliplatin response. These results demonstrate, that the ability to assess anti-tumor activity in well-characterized xenografts in correlation with particular genetic or molecular characteristics may support the development of new therapeutic regimens. Conclusions from what we discussed are: drug discovery, systems biology, and translational research are moving together to address all the new hallmarks of cancer - increasing the success rate of drug discovery panels of patient derived xenograft models represent an important tool for translational research to address clinically relevant questions in a standardized and strictly controlled fashion predictive value of the preclinical models is increasing steadily - they show a high concordance with the clinical specimens concerning marker expression and response to therapy Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr A14.

Abstract 1584: Characterization of 28 patient-derived colorectal cancer xenografts and analysis for predictive response markers to current standard treatment

Proceedings: AACR 102nd Annual Meeting 2011‐‐ Apr 2‐6, 2011; Orlando, FLnnImproved methods for the preclinical characterization of anti-cancer drugs in larger panels of primary lung cancer xenograft models have recently described (Fichtner et al. Clin Cancer Res. 2008;14(20):6456-68). These models consider more thoroughly the heterogeneity and genetic instability of tumors in a way ultimately enabeling treatment decisions on an individualized basis.nnThe approach is based on large scale screening for activity in new in vivo models combined with extensive characterization of these models (gene expression and mutations). Bioinformatic methods are then used to further evaluate the mechanism of action of new drugs, to investigate possible resistance mechanisms and provide a rationale for combination therapies.nnHere we describe the development of a panel of primary patient derived colorectal cancer models. Primary colon carcinoma tissue samples were collected using a standardised procedure. Tumour pieces were transplanted onto immunodeficient mice immediately after surgery. A panel of 28 stably passagable colon cancer xenografts could be established as permanent tumour models. The results obtained so far, show that these patient-derived colon cancer models feature a high coincidence with the original tumour regarding histology and gene expression profiling. Response analysis according to clinical trial criteria revealed that the standard drug Oxaliplatin induce a overall response in 8 of the 28 xenograft models.nnGene expression analyses were performed to identify predictive profiles for Oxaliplatin sensitivity. As one example, the effectiveness of Oxaliplatin seems to be related to the differential expression of genes which products are involved in the wnt pathway.nnThe analysis of drug sensitivity in this set of models together with bioinformatics coorelations will provide hypotheses for further studies to identify patients that will more likely benefit from a drug as well as provide a rationale for combinations to be tested in clinical trials.nnThese new approaches may help to better integrate research results in clinical development and may increase the success of new anti-cancer drugs in clinical studies.nnCitation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 1584. doi:10.1158/1538-7445.AM2011-1584