@phdthesis{Landwehr2023,
  author    = {Landwehr, Laura-Sophie},
  title     = {Steroid Hormones and Cancer Immunity - learning from Adrenocortical Carcinoma},
  doi       = {10.25972/OPUS-25189},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-251895},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Adrenocortical carcinoma (ACC) is a rare, but highly aggressive endocrine malignancy. Tumor-related hypercortisolism is present in 60 \% of patients and associated with worse outcome. While cancer immunotherapies have revolutionized the treatment of many cancer entities, the results of initial studies of different immune checkpoint inhibitors in ACC were heterogeneous. Up to now, five small clinical trials with a total of 121 patients have been published and demonstrated an objective response in only 17 patients. However, one of the studies, by Raj et al., reported a clinically meaningful disease control rate of 52 \% and a median overall survival of almost 25 months suggesting that a subgroup of ACC patients may benefit from immunotherapeutic approaches. Following the hypothesis that some ACCs are characterized by a glucocorticoid-induced T lymphocytes depletion, several studies were performed as part of the presented thesis. First, the immune cell infiltration in a large cohort of 146 ACC specimens was investigated. It was demonstrated for the first time, and against the common assumption, that ACCs were infiltrated not only by FoxP3+ regulatory T cells (49.3 \%), but also that a vast majority of tumor samples was infiltrated by CD4+ TH cells (74 \%) and CD8+ cytotoxic T cells (84.3 \%), albeit the immune cell number varied heterogeneously and was rather low (median: 7.7 CD3+ T cells / high power field, range: 0.1-376). Moreover, the presence of CD3+-, CD4+- and CD8+ ACC-infiltrating lymphocytes was associated with an improved recurrence-free (HR: 0.31 95 \% CI 0.11-0.82) and overall survival (HR: 0.47 96 \% CI 0.25-0.87). Particularly, patients with tumor-infiltrating CD4+ TH cells without glucocorticoid excess had a significantly longer overall survival compared to patients with T cell-depleted ACC and hypercortisolism (121 vs. 27 months, p = 0.004). Hence, the impact of glucocorticoids might to some extent be responsible for the modest immunogenicity in ACC as hypercortisolism was reversely correlated with the number of CD4+ TH cells. Accordingly, CD3+ T cells co-cultured with steroidogenic NCI-H295R ACC cells demonstrated in vitro an enhanced anti-tumoral cytotoxicity by secreting 747.96 ±225.53 pg/ml IFN-γ in a therapeutically hormone-depleted microenvironment (by incubation with metyrapone), versus only 276.02 ±117.46 pg/ml IFN-γ in a standard environment with glucocorticoid excess. Other potential biomarkers to predict response to immunotherapies are the immunomodulatory checkpoint molecules, programmed cell death 1 (PD-1) and its ligand PD-L1, since both are targets of antibodies used therapeutically in different cancer entities. In a subcohort of 129 ACCs, expressions of both molecules were heterogeneous (PD-1 17.4 \%, range 1-15; PD-L1 24.4 \%, range 1 - 90) and rather low. Interestingly, PD-1 expression significantly influenced ACC patients´ overall (HR: 0.21 95 \% CI 0.53-0.84) and progression- free survival (HR: 0.30 95 \% CI 0.13-0.72) independently of established factors, like ENSAT tumor stage, resection status, Ki67 proliferation index and glucocorticoid excess, while PD-L1 had no impact. In conclusion, this study provides several potential explanations for the heterogeneous results of the immune checkpoint therapy in advanced ACC. In addition, the establishment of PD-1 as prognostic marker can be easily applied in routine clinical care, because it is nowadays anyway part of a detailed histo-pathological work-up. Furthermore, these results provide the rationale and will pave the way towards a combination therapy using immune checkpoint inhibitors as well as glucocorticoid blockers. This will increase the likelihood of re-activating the immunological anti-tumor potential in ACC. However, this will have to be demonstrated by additional preclinical in vivo experiments and finally in clinical trials with patients.},
  subject      = {Steroidhormon},
  language  = {en}
}
@phdthesis{Marquardt2023,
  author    = {Marquardt, Andr{\´e}},
  title     = {Machine-Learning-Based Identification of Tumor Entities, Tumor Subgroups, and Therapy Options},
  doi       = {10.25972/OPUS-32954},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-329548},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Molecular genetic analyses, such as mutation analyses, are becoming increasingly important in the tumor field, especially in the context of therapy stratification. The identification of the underlying tumor entity is crucial, but can sometimes be difficult, for example in the case of metastases or the so-called Cancer of Unknown Primary (CUP) syndrome. In recent years, methylome and transcriptome utilizing machine learning (ML) approaches have been developed to enable fast and reliable tumor and tumor subtype identification. However, so far only methylome analysis have become widely used in routine diagnostics. The present work addresses the utility of publicly available RNA-sequencing data to determine the underlying tumor entity, possible subgroups, and potential therapy options. Identification of these by ML - in particular random forest (RF) models - was the first task. The results with test accuracies of up to 99\% provided new, previously unknown insights into the trained models and the corresponding entity prediction. Reducing the input data to the top 100 mRNA transcripts resulted in a minimal loss of prediction quality and could potentially enable application in clinical or real-world settings. By introducing the ratios of these top 100 genes to each other as a new database for RF models, a novel method was developed enabling the use of trained RF models on data from other sources. Further analysis of the transcriptomic differences of metastatic samples by visual clustering showed that there were no differences specific for the site of metastasis. Similarly, no distinct clusters were detectable when investigating primary tumors and metastases of cutaneous skin melanoma (SKCM). Subsequently, more than half of the validation datasets had a prediction accuracy of at least 80\%, with many datasets even achieving a prediction accuracy of - or close to - 100\%. To investigate the applicability of the used methods for subgroup identification, the TCGA-KIPAN dataset, consisting of the three major kidney cancer subgroups, was used. The results revealed a new, previously unknown subgroup consisting of all histopathological groups with clinically relevant characteristics, such as significantly different survival. Based on significant differences in gene expression, potential therapeutic options of the identified subgroup could be proposed. Concludingly, in exploring the potential applicability of RNA-sequencing data as a basis for therapy prediction, it was shown that this type of data is suitable to predict entities as well as subgroups with high accuracy. Clinical relevance was also demonstrated for a novel subgroup in renal cell carcinoma. The reduction of the number of genes required for entity prediction to 100 genes, enables panel sequencing and thus demonstrates potential applicability in a real-life setting.},
  subject      = {Maschinelles Lernen},
  language  = {en}
}