Eva Oswald

Capturing complex tumour biology in vitro: histological and molecular characterisation of precision cut slices

Precision-cut slices of in vivo tumours permit interrogation in vitro of heterogeneous cells from solid tumours together with their native microenvironment. They offer a low throughput but high content in vitro experimental platform. Using mouse models as surrogates for three common human solid tumours, we describe a standardised workflow for systematic comparison of tumour slice cultivation methods and a tissue microarray-based method to archive them. Cultivated slices were compared to their in vivo source tissue using immunohistochemical and transcriptional biomarkers, particularly of cellular stress. Mechanical slicing induced minimal stress. Cultivation of tumour slices required organotypic support materials and atmospheric oxygen for maintenance of integrity and was associated with significant temporal and loco-regional changes in protein expression, for example HIF-1α. We recommend adherence to the robust workflow described, with recognition of temporal-spatial changes in protein expression before interrogation of tumour slices by pharmacological or other means.

A high-content image analysis approach for quantitative measurements of chemosensitivity in patient-derived tumor microtissues

Organotypic, three-dimensional (3D) cancer models have enabled investigations of complex microtissues in increasingly realistic conditions. However, a drawback of these advanced models remains the poor biological relevance of cancer cell lines, while higher clinical significance would be obtainable with patient-derived cell cultures. Here, we describe the generation and data analysis of 3D microtissue models from patient-derived xenografts (PDX) of non-small cell lung carcinoma (NSCLC). Standard of care anti-cancer drugs were applied and the altered multicellular morphologies were captured by confocal microscopy, followed by automated image analyses to quantitatively measure phenotypic features for high-content chemosensitivity tests. The obtained image data were thresholded using a local entropy filter after which the image foreground was split into local regions, for a supervised classification into tumor or fibroblast cell types. Robust statistical methods were applied to evaluate treatment effects on growth and morphology. Both novel and existing computational approaches were compared at each step, while prioritizing high experimental throughput. Docetaxel was found to be the most effective drug that blocked both tumor growth and invasion. These effects were also validated in PDX tumors in vivo. Our research opens new avenues for high-content drug screening based on patient-derived cell cultures, and for personalized chemosensitivity testing.

Context-Based Normalization of Histological Stains Using Deep Convolutional Features

While human observers are able to cope with variations in color and appearance of histological stains, digital pathology algorithms commonly require a well-normalized setting to achieve peak performance, especially when a limited amount of labeled data is available. This work provides a fully automated, end-to-end learning-based setup for normalizing histological stains, which considers the texture context of the tissue. We introduce Feature Aware Normalization, which extends the framework of batch normalization in combination with gating elements from Long Short-Term Memory units for normalization among different spatial regions of interest. By incorporating a pretrained deep neural network as a feature extractor steering a pixelwise processing pipeline, we achieve excellent normalization results and ensure a consistent representation of color and texture. The evaluation comprises a comparison of color histogram deviations, structural similarity and measures the color volume obtained by the different methods.

Abstract 4815: Humanized single mouse trial: A preclinical platform feasible for immune-oncology drug screening and translational biomarker development

The field of cancer immunology is rapidly moving towards innovative therapeutic strategies. As a consequence the need for robust and predictive preclinical platforms arises just as well. The current project aims to establish a drug screening workflow bridging between innovative mouse models and clinical biomarker development. A total of 69 NOG (NOD/Shi-scid/IL-2Rγnull) mice were engrafted with CD34+ hematopoietic stem cells. Thereafter, tumor material from 11 different lung cancer patient derived xenograft models (NSCLC PDX) was implanted subcutaneously. Individual mice were treated with α-CTLA-4, α-PD-1 or the combination thereof. With n=1 per treatment arm and model the study design followed the screening approach of the single mouse trial (SMT). Infiltration of human immune cells was detected by flow cytometry (FC) and immunohistochemistry (IHC) in hematopoietic organs and tumor tissue. A computerized analysis for digitized whole-slide images of the samples was used to quantify the lymphocyte infiltration using color classification and morphological image processing techniques. All 3 treatment arms displayed a discrete activity pattern throughout the PDX panel. Tumor models with high tumor infiltrating lymphocyte (TIL) rates in the donor patient material tended to be more sensitive towards checkpoint inhibitor treatment as models with low rates. Numbers of TILs in the PDX detected by FC and IHC were significantly increased in the treatment groups as compared to control vehicle. In parallel, hematopoietic organs showed high (>25%) amounts of huCD45 cells in all groups and models. PDX models being sensitive towards checkpoint inhibitor treatment (responders) displayed a higher percentage of DAB+ nuclei in huCD45 IHC stains than non-responder models as determined by image analysis. Irrespective thereof, in responders as well as non-responders the treatment with checkpoint inhibitors enhanced the percentage of DAB+ nuclei. Whole-slide image analysis of the HE 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4815. doi:10.1158/1538-7445.AM2017-4815

Abstract 590: Co-injection of human monocytes improves the in vivo antitumoral activity of bevacizumab in two NSCLC PDX models

Nowadays, an increasing number of monoclonal antibodies (mAbs) that specifically target malignant cells or interfere with different compartments of the tumor microenvironment are available for cancer therapy. They take effect via different modes of action including the initiation of a tumor-targeting immune response. Preclinical platforms such as PDX models have to be improved to better recapitulate all possible modes of action for mAbs as well as other immune-modulating agents. In the current study, we evaluated the antitumoral activity of Bevacizumab in two NSCLC PDX growing subcutaneously in NMRI nu/nu mice with and without co-injection of human monocytes. Two NSCLC PDX, LXFA 2478 and LXFA 677, were subcutaneously implanted into 4-6 weeks old female NMRI nu/nu mice (Harlan, Denmark). Tumor models were chosen based on their VEGFA expression level. Additionally, LXFA 2478 and LXFA 677 show high expression of TLR2 and CD14 respectively, two factors known to be involved in monocyte attraction. When median tumor size reached 150 - 300 mm3, mice were equally distributed to treatment groups (n = 4/group). Animals were treated once weekly for 7 cycles with a) control vehicle b) control vehicle + 5×106 human monocytes c) Bevacizumab at 40 mg/kg/d and d) Bevacizumab + 5×106 human monocytes. Tumor volume was determined twice weekly by caliper measurement. At the end of the study, tumor and lymphatic organs of the animals were harvested and subsequent IHC analysis for human CD14, CD68 and CD163 was performed. In both investigated tumor models, Bevacizumab showed moderate antitumoral activity with a maximal tumor load reduction of 71% (LXFA 2478) and 84% (LXFA677) as compared to untreated controls on days 39 and 35. The co-injection of human monocytes markedly enhanced the therapeutic effect of Bevacizumab in both NSCLC PDX: maximal tumor load reduction was 89% in LXFA 2478 and 95% in LXFA 677 in combination groups. The injection of monocytes alone did not affect tumor growth as compared to untreated control. As predicted by the high VEGFA expression, NSCLC PDX showed high sensitivity against Bevacizumab. This antitumoral activity was increased by 18% and 11% for LXFA 2478 and LXFA 677, respectively, through the additional injection of monocytes. Monocyte recruiting factors (namely CD14 and TLR2) likely contribute to the mechanism of action. Monocytes and macrophages have been reported to induce antibody-dependent cytotoxicity and phagocytosis of tumor cells in the presence of IgG anti-tumor mAbs, like Bevacizumab. Our results confirm these observations in a PDX based NSCLC in vivo model. Therefore, the study highlights the suitability of PDX for immuno-oncology approaches by supplementation of the murine host with human immune cells. This advanced PDX approach will lead to more predictive preclinical data for innovative mAb′s and other compounds acting via the activation of monocytes and related immune cells. Citation Format: Cordula Tschuch, Kerstin Klingner, Anne Lohr, Yana Raeva, Teppo Haapaniemi, Eva Oswald, Julia B. Schuler. Co-injection of human monocytes improves the in vivo antitumoral activity of bevacizumab in two NSCLC PDX models. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 590.

Abstract 630: Co-injection of human fibroblasts significantly enhances tumorigenicity of orthotopically implanted human non-small cell lung cancer cells in immunocompromised mice

Human tumor xenografts in immunodeficient mice have led to valuable insights into the biology of human cancer. The corresponding limitations and pitfalls of xenograft models have been described extensively. To tackle some limitations, we supplemented the murine tumor microenvironment (TME) with human stromal fibroblasts to mirror tumor-stroma cross-talk in vivo. Non-small cell lung cancer (NSCLC) cell lines Calu-1 and H1437 were co-injected with different types of fibroblasts into NOG (NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ) mice in 3 independent experiments (n = 4-6/group; 106 in total). Human Dermal Fibroblasts (HDFs), lung cancer associated fibroblasts (CAFs), corresponding normal fibroblasts (NFs) and NSCLC cells were implanted into the lungs of 4-6 week old recipient mice. The influence of extracellular matrix components was evaluated by using either a matrigel/collagen mixture or alginate capsules. Tumor load was determined via overall survival (OS) and flow cytometry (FC). When mice had to be sacrificed, organs were collected and analyzed by FC and patho-histological examination. In a subsequent experiment, mice were sacrificed and tumors analyzed at defined time points and similar analyses performed. OS and metastatic pattern were similar in both cell lines when injected in mono-culture. OS of Calu-1 injected mice was not influenced when cells were co-injected with HDF but prolonged when co-injected with CAFs or NFs. In H1437 bearing mice, OS was significantly reduced when cells were co-injected with HDF or NF (Log-rank [Mantel-cox] test; p Our results demonstrate the major impact of fibroblasts on tumor cell behavior in a preclinical setting. With the successful co-cultivation of human fibroblast and NSCLC cells in vivo, it will be possible to study tumor-stroma interactions in a clinically relevant mouse model. Once validated by a compound screening approach, this model will help to reduce drug failure rates and contribute to a more efficient development of urgently needed novel anti-cancer treatments. Citation Format: Eva Oswald, Vitor E. Santo, Albin Rudisch, Catarina Brito, Wolfgang Sommergruber, Helmut Dolznig, Julia B. Schuler. Co-injection of human fibroblasts significantly enhances tumorigenicity of orthotopically implanted human non-small cell lung cancer cells in immunocompromised mice. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 630.

Abstract A10: Establishment and characterization of a patient-derived non-small cell lung cancer mouse model of acquired resistance towards anti-EGFR treatment

Non-small cell lung cancer (NSCLC) is the largest subgroup of lung cancer, occurring at a frequency of over 80% of lung cancer cases. In up to 30% of NSCLC patients the oncogenic driver of tumor growth is a constitutively activated EGF receptor (EGFR), which plays a critical role in regulating multiple cellular processes, including proliferation, survival and apoptosis. Although these patients gain great benefit from treatment with EGFR tyrosine kinase inhibitors (TKI, e.g. erlotinib or gefitinib), development of resistance is inevitable. To model the emergence of drug resistance, an EGFR driven, gefitinib sensitive, patient-derived xenograft (PDX) NSCLC model was treated continuously with gefitinib in immunocompromised mice. The dose of daily treatment was adjusted according to tumor growth over a period of up to 91 days. In a first phase, dosing was high (40-50 mg/kg) to eradicate EGFR TKI sensitive cells. At the time point of maximal antitumoral activity dosing was reduced to 20-30 mg/kg ( = low dose) to preserve selection pressure. Between 69 and 91 days after dosing was initiated, drug-resistant tumors emerged in 4 out of 10 mice under high dose treatment. Resistant tumor fragments, which were re-implanted into a new cohort of mice and continuously treated with gefitinib, kept resistance also under high dose treatment. A comprehensive analysis using Western blot (WB), qPCR and sequencing was performed to identify the reason for resistance. In WB analysis we could show that signalling through EGFR was completely abrogated in all four resistant tumor sublines. Neither secondary mutations in EGFR (ex19-21) or KRAS (ex 1+2) could be detected, nor was the expression of cMET, AXL, HGF, PTEN or HER3 significantly increased in resistant tumors as shown by sequencing and qPCR respectively. However a more comprehensive WB analysis revealed several genes being activated in resistant compared to primary tumors. Depending on the subline, phospho-(p)-cMET, p-AKT, AXL & p-AXL, p-cRAF, p-MEK, p-ERK as well as HER3, IGF-R and ALK were up-regulated in resistant tumors. Based on these data we determined that signalling pathways such as the (RAS)-cRAF-MEK-ERK signalling cascade or enhanced cMET/AKT signalling were activated in drug-resistant tumors. These signalling pathways are known to be alternative pathways for EGFR signalling also in patients with acquired resistance indicating the clinical relevance of these models. To shed more light into the mechanism of resistance, whole exome sequencing analyses of the four resistant sublines as well as the original tumor model are underway. In summary, we have developed four NSCLC tumor sublines each harbouring a different mechanism of resistance to EGFR TKI treatment, modelling the emergence of drug-resistant NSCLC in patients. Herewith a strong preclinical tool for the development of innovative compounds targeting acquired EGFR resistance is now available. Citation Format: Cordula Tschuch, Kerstin Klingner, Dorothee Lehnhard, Anne Lohr, Yana Raeva, Anne-Lise Peille, Eva Oswald, Julia B. Schuler. Establishment and characterization of a patient-derived non-small cell lung cancer mouse model of acquired resistance towards anti-EGFR treatment. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr A10.

Abstract 5023: NSCLC PDX model for the evaluation of immuno-oncological treatment strategies

Patient-derived tumor xenografts (PDX) have played a major role in the development of new cancer therapies and their strengths and weaknesses have gradually been elucidated. One major drawback of PDX is the lack of an immunological competent host. To overcome this hurdle we supplemented NSG/NOG mice with human hematopoietic stem cells (HSC) and subsequently examined growth characteristics of the non-small cell lung cancer (NSCLC) PDX model LXFA 923 in these mice. In parallel we monitored the presence of human and murine immune cells in different organs of the mouse. HSC (2×106) cells were isolated from healthy donors and injected intravenously into sub-lethally irradiated NSG or NOG mice (n = 43 mice in 3 ind. exp.). After 8 weeks LXFA 923 was transplanted subcutaneously (s.c.) into the pretreated mice. Murine peripheral blood was examined by flow cytometry for common murine and human markers expressed on immune cells (hCD14, mCD14, hCD3, mCD3, hCD56, mCD56, hCD19) once weekly. At the end of the experiment tumors and organs were analyzed for human cancer (CD44, CD133, CDCP1, CD166, CD24) and immune cell markers (hCD14, hCD3, hCD56, hCD19) by flow cytometry. Tumors and organs were additionally histologically and immunohistochemically examined. Growth of the implanted tumors was monitored by caliper measurement. Mice bearing only the subcutaneous PDX or the HSC served as control groups. Stable engraftment of human immune cells in immune-compromized mice was successfully achieved. Human immune cells expressing T-, B-, NK- and stem cell markers could be detected in different compartments (bone marrow, peripheral blood and spleen) of the tumor-bearing as well as non-tumor bearing mice. Furthermore, infiltrates of human monocytes (CD14+) as well as T cells (CD3+) could be detected in s.c. implanted tumor tissue. Implantation of LXFA 923 did not influence the proliferation of human immune cells in recipient mice. Growth behavior of the s.c. implanted PDX was not affected by the engraftment of HSC in the murine host. The histological architecture of LXFA 923 was similar when implanted s.c. in humanized or immunodeficient mice and it still closely resembles the patient donor material. In conclusion, our investigations validate the analysis of PDX in mice engrafted with human immune cells, as it enables the interaction of tumor cells with human immune cells as well as with murine stroma to be investigated. This preclinical PDX based in vivo platform provides a further step to support the development of new drugs targeting the host immune response. Citation Format: Eva Oswald, Kerstin Klingner, Dorothee Lenhard, Gabriele Niedermann, Julia B. Schuler. NSCLC PDX model for the evaluation of immuno-oncological treatment strategies. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5023. doi:10.1158/1538-7445.AM2015-5023

Abstract 3926: Influence of tumor microenvironment on engraftment capacity of hematological cell lines in immunocompromized mice

Hematological malignancies account for about 10% of newly diagnosed cancers in the US, their heterogeneity and diverse biological characteristics present unique therapeutic challenges. In order to develop more promising therapeutic strategies, the establishment of functional and reproducible in vivo models is widely pursued. A panel of 18 hematological cell lines comprising different entities including leukemia (acute lymphoblastic AML; chronic myeloid CML; chronic lymphatic leukemia CLL), diffuse large B cell lymphoma (DLBCL), as well as multiple myeloma (MM) was evaluated for in vivo engraftment using different mouse strains and application routes. 18 different cell lines were inoculated into 4 different immuncompromized mouse strains (NOG; NSG; NOD/SCID; SCID/beige nude) in 35 independent experiments. Cells were injected intravenously (i.v.), intratibialy (i.t.) or intraperitonealy (i.p.). Tumor growth was monitored via a) determination of overall survival, b) fluorescence-based in vivo imaging (IVI, Bruker FX, using CF750 labeled anti-hu CD33, CD19, CD45) c) flow cytometry and d) histological and IHC examination. Engraftment of hematological cell lines was clearly influenced by tumor entity, mouse strain and application route. The investigated AML cell lines engrafted in all three applied mouse strains (NOG; NSG; NOD/SCID) and injection routes. MOLM-13 infiltrated the bone marrow (BM) and the spleen (>90% infiltration, 21 days after tumor cell injection (=d21)), whereas MV4-11 cells disseminated mainly to the spleen (5% BM and 15% spleen infiltration, d21). In contrast THP-1 (80% BM and 2% spleen infiltration, d21) as well as NOMO-1 cells (30% BM and 3% spleen infiltration, d21) grew predominantly in the murine BM. MM cell lines (L363, RPMI8826, MM1R, MM1S) as well as B-CLL cell lines (Mec-1) failed to engraft reliably in NOD/SCID mice, irrespective of the application route. MM cells engrafted predominantly in the BM (20 - 60% depending on the cell line) of NSG mice when injected i.t.. i.v. injected MM cells did engraft in the BM but to a significantly lower extent than when injected i.t. (e.g. 20% vs. 2% BM infiltration of RPMI8226 cells, d35). Mec-1 cells infiltrated the BM and spleen of NSG and NOG mice reliably when injected i.v. and i.p. in the same individual (60% BM and 15% spleen infiltration, d28), whereas the same approach did not result in any engraftment when using SCID/beige nude mice. 3 out of 4 investigated DLBCL cell lines engrafted in vivo. Disseminated growth could be observed in NOG mice when injecting the cells i.v. or i.t. In all cases, tumors grew faster and infiltrated the BM more aggressively when injected I.t.. In summary, the engraftment of hematological cell lines strongly depends on the tumor microenvironment. The direct contact to the BM niche is a major survival benefit for the tumor cells as is the lack of NK cells in NOG and NSG mice. Citation Format: Eva Oswald, Kerstin Klingner, Ralph Waesch, Katja Zirlik, Monika Engelhardt, Julia B. Schuler. Influence of tumor microenvironment on engraftment capacity of hematological cell lines in immunocompromized mice. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3926. doi:10.1158/1538-7445.AM2014-3926