Arbeitsgruppe Becker

Translational Oncology Skin Cancer Research

Prof. Dr. Dr. Jürgen Becker

High-definition spatial transcriptomics (HD ST) analysis of tumor epidermotropisms. (A) H&E staining of a FFPE skin cancer tissue. (B) Spatial RNA sequencing using Visium Spatial HD technology recovered the histological structures and cellular heterogeneities of the tissue in high precision. (C) Deep learning approaches employing convolutional neural network were applied to the HD ST data to generate high-confidence cell sets for downstream analysis. (D) Cell trajectory was constructed to demonstrate the continual heterogeneities of tumor cells according to spatial localization, i.e., epidermal (purple), intermediate (red), core (orange). © TSCR

In skin cancer, translational cancer research is facilitated by the fact that most tumors are diagnosed at an early stage, have a staged course of disease (i.e. locoregional metastases preceding distant metastases), and even in advanced disease skin lesions often coexist with visceral metastases. These characteristics enable a detailed examination of tumor tissues using state-of-the-art methods, which provides insights not only into the tumor cells themselves, but also the tumor microenvironment (TME), particularly the immune system.
Indeed, cancer is not merely an accumulation of cancer cells; it involves a complex interplay between stromal, immune, and neoplastic cells. This interaction is crucial not only for the disease's natural progression but also for its response to therapy.

Until recently, this complexity could only be functionally investigated using model systems such as patient-derived mouse models, CAM models, tumor organoids, or tumor-on-a-chip technology. These models aim to map the reciprocal effects of interactions within the TME and relevant tumor properties such as invasion, angiogenesis, and immune responses. However, even these advanced model systems remain artificial and do not fully reflect the conditions within the patient.
Hence it is exciting that technical advances such as single cell multiomics (scRNA-seq, scATAC-seq, scM&T-seq, REAP-seq), spatial trancriptomics and phenotomics together with advanced integrative analysis methods such as Joint Latent Variable and Bayesian Models now allow comprehensive ex vivo analyses of patient-derived tumor tissue to gain functional insights. This applies in particular if the material to be examined is obtained under defined conditions, i.e. in specifically conducted low-interventional clinical trials (LIT) as defined in the EU Clinical Trials Regulation No 536/2014 (e.g. ‘FieldCancerization’ or ‘CemiFirst’), or in translational research programs accompanying clinical trials (e.g. ‘IMMUNED’ or ‘ADMEC’). Moreover, recent technical developments now enable the analysis of FFPE (formalin-fixed, paraffin-embedded) tissue, allowing for the detailed examination of archived tissue that has been prospectively collected in tissue registries (e.g. ‘MCC TRIM’).
Using these approaches we were able to identify not only novel surrogate (tumor burden, early diagnosis), prognostic (spontaneous clinical course) and predictive (response to therapy) biomarkers for skin cancer, but also new therapeutic strategies. For example, the combined inhibition of DNA repair and DNA damage response or the induction of neuroblastic transformation of MCC cells.

Moreover, our findings suggest that applying immunotherapy at an earlier stage, actually even before the adjuvant setting, could significantly improve clinical outcomes. This approach will soon be evaluated in an upcoming clinical trial, which will be complemented by a robust translational research program.

Future projects and goals
Over tte last few years, we have succeeded in establishing a prospective collection of clinically annotated biosamples of skin tumors, some of which are even taken sequentially under defined conditions and reflect the dynamic changes in the tumor and its microenvironment. These samples will now be used to answer specific questions, e.g. the spatial distribution of tumor heterogeneity, the influence of the activation of endogenous retroviruses on immunotherapies, or the role of the lesional microbiome on tumor progression. The above-mentioned studies are not only very expensive and time-consuming, but also data-intensive. It is therefore planned to carry out some of these on tissue micro-arrays (TMAs), whereby a 3D overview of the tissue structure will first be established from the tissue blocks using micro-CT. This approach not only enables ideal positioning of the tissue cores, but also allows to project the findings of the multi-omic analyses back into the 3D tissue context.