Felix C. Popp

Background/Objectives: Despite the introduction of targeted therapies such as Nivolumab, survival outcomes for patients with esophageal adenocarcinoma remain poor. During tumorigenesis, some tumors develop auxotrophy by downregulation of Argininosuccinate Synthetase-1 (ASS1), making them reliant on external arginine supply and thus potentially susceptible to arginine deprivation therapy. Arginine deprivation therapy with agents such as pegargiminase has shown improved survival in patients with pleural mesothelioma exhibiting ASS1 loss in tumor cells. Therefore, we investigated the prevalence of ASS1 loss in esophageal adenocarcinoma. Methods: First, we compared the staining patterns of three antibodies for ASS1 with RNA in situ Scope analysis results to identify the most reliable antibody for ASS1 immunohistochemical staining in esophageal adenocarcinoma. Subsequently, we performed ASS1 immunohistochemical staining on samples from 97 patients who underwent curative resection. The staining results were classified into three categories based on expression levels: negative, low-positive, and positive. Results: Among all included patients, 6.2% exhibited an ASS1 loss, and 6.2% showed low ASS1 expression. Notably, patients with an ASS1 loss did not demonstrate a response to neoadjuvant therapy. Patients with ASS1 loss or low expression were significantly younger. Conclusions: Our findings indicate that approximately 12.4% of patients with esophageal adenocarcinoma may be eligible and could potentially benefit from arginine deprivation therapy. This underscores the urgent need for clinical trials evaluating the efficacy of pegargiminase in this patient population. Additionally, incorporating ASS1 immunohistochemical staining into pre-neoadjuvant biopsy assessments should be considered to optimize neoadjuvant treatment strategies and advance the implementation of personalized cancer therapy.

Background: The aim of this study was to establish a deep learning prediction model for neoadjuvant FLOT chemotherapy response. The neural network utilized clinical data and visual information from whole-slide images (WSIs) of therapy-naïve gastroesophageal cancer biopsies. Methods: This study included 78 patients from the University Hospital of Cologne and 59 patients from the University Hospital of Heidelberg used as external validation. Results: After surgical resection, 33 patients from Cologne (42.3%) were ypN0 and 45 patients (57.7%) were ypN+, while 23 patients from Heidelberg (39.0%) were ypN0 and 36 patients (61.0%) were ypN+ (p = 0.695). The neural network had an accuracy of 92.1% to predict lymph node metastasis and the area under the curve (AUC) was 0.726. A total of 43 patients from Cologne (55.1%) had less than 50% residual vital tumor (RVT) compared to 34 patients from Heidelberg (57.6%, p = 0.955). The model was able to predict tumor regression with an error of ±14.1% and an AUC of 0.648. Conclusions: This study demonstrates that visual features extracted by deep learning from therapy-naïve biopsies of gastroesophageal adenocarcinomas correlate with positive lymph nodes and tumor regression. The results will be confirmed in prospective studies to achieve early allocation of patients to the most promising treatment.

Purpose: Patients with pancreatic ductal adenocarcinoma (PDAC) have yet to experience significant benefits from targeted therapy. Olaparib is currently the only active substance in BRCA-mutated PDACs that successfully influences the DNA repair of carcinoma cells. H2AX belongs to the histone family and is known as a part of the DNA repair system. The inhibition of γ-H2AX could lead to the inhibition of mitotically active tumor cells. Therefore, we aimed to evaluate the predictive value of the γ-H2AX in patients with PDAC. Methods: All included patients (n = 311) received a pancreatic resection with curative intention in one of our PANCALYZE study centers. Subsequently, they were enrolled in a standardized follow-up protocol. Immunohistochemical stainings for γ-H2AX were conducted on tissue microarrays. Results: Patients exhibiting high levels of γ-H2AX expression experience more frequent R1 resections, indicating advanced tumor stages in this subgroup. Additionally, patients with high γ-H2AX expression demonstrated significantly poorer survival compared to those with low expression (median OS: 15 vs. 25 months, p < 0.001). In multivariate analyses, high γ-H2AX expression could be identified as an independent risk factor for worse patient survival. Moreover, high γ-H2AX expression could be more frequently observed in the more aggressive basal-like subtype. Conclusion: γ-H2AX can be characterized as a predictive biomarker for poorer patient survival. Consequently, upcoming clinical trials focused on the efficacy of targeted therapies influencing the DNA repair system and radiotherapy should evaluate γ-H2AX as a potential biomarker for therapy response. Furthermore, γ-H2AX may serve as a viable target for treatment in the future.

AIM: In this study, we use Artificial Intelligence (AI), including Machine (ML) and Deep Learning (DL), to predict the long-term survival of resectable esophageal cancer (EC) patients in a high-volume surgical center. Our objective is to evaluate the predictive efficacy of AI methods for survival prognosis across different time points of oncological treatment. This involves comparing models trained with clinical data, integrating either Tumor, Node, Metastasis (TNM) classification or tumor biomarker analysis, for long-term survival predictions. METHODS: In this retrospective study, 1002 patients diagnosed with EC between 1996 and 2021 were analyzed. The original dataset comprised 55 pre- and postoperative patient characteristics and 55 immunohistochemically evaluated biomarkers following surgical intervention. To predict the five-year survival status, four AI methods (Random Forest RF, XG Boost XG, Artificial Neural Network ANN, TabNet TN) and Logistic Regression (LR) were employed. The models were trained using three predefined subsets of the training dataset as follows: (I) the baseline dataset (BL) consisting of pre-, intra-, and postoperative data, including the TNM but excluding tumor biomarkers, (II) clinical data accessible at the time of the initial diagnostic workup (primary staging dataset, PS), and (III) the PS dataset including tumor biomarkers from tissue microarrays (PS + biomarkers), excluding TNM status. We used permutation feature importance for feature selection to identify only important variables for AI-driven reduced datasets and subsequent model retraining. RESULTS: Model training on the BL dataset demonstrated similar predictive performances for all models (Accuracy, ACC: 0.73/0.74/0.76/0.75/0.73; AUC: 0.78/0.82/0.83/0.80/0.79 RF/XG/ANN/TN/LR, respectively). The predictive performance and generalizability declined when the models were trained with the PS dataset. Surprisingly, the inclusion of biomarkers in the PS dataset for model training led to improved predictions (PS dataset vs. PS dataset + biomarkers; ACC: 0.70 vs. 0.77/0.73 vs. 0.79/0.71 vs. 0.75/0.69 vs. 0.72/0.63 vs. 0.66; AUC: 0.77 vs. 0.83/0.80 vs. 0.85/0.76 vs. 0.86/0.70 vs. 0.76/0.70 vs. 0.69 RF/XG/ANN/TN/LR, respectively). The AI models outperformed LR when trained with the PS datasets. The important features shared after AI-driven feature selection in all models trained with the BL dataset included histopathological lymph node status (pN), histopathological tumor size (pT), clinical tumor size (cT), age at the time of surgery, and postoperative tracheostomy. Following training with the PS dataset with biomarkers, the important predictive features included patient age at the time of surgery, TP-53 gene mutation, Mesothelin expression, thymidine phosphorylase (TYMP) expression, NANOG homebox protein expression, and indoleamine 2,3-dioxygenase (IDO) expressed on tumor-infiltrating lymphocytes, as well as tumor-infiltrating Mast- and Natural killer cells. CONCLUSION: Different AI methods similarly predict the long-term survival status of patients with EC and outperform LR, the state-of-the-art classification model. Survival status can be predicted with similar predictive performance with patient data at an early stage of treatment when utilizing additional biomarker analysis. This suggests that individual survival predictions can be made early in cancer treatment by utilizing biomarkers, reducing the necessity for the pathological TNM status post-surgery. This study identifies important features for survival predictions that vary depending on the timing of oncological treatment.

Purpose: The pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal cancer entities. Effective therapy options are still lacking. The tumor microenvironment possibly bears further treatment possibilities. This study aimed to describe the expression patterns of four established carcinoma-associated fibroblast (CAFs) markers and their correlation in PDAC tissue samples. Methods: This project included 321 patients with PDAC who underwent surgery with a curative intent in one of the PANCALYZE study centers. Immunohistochemical stainings for FAP, PDGFR, periostin, and SMA were performed. The expression patterns of each marker were divided into low- and high-expressing CAFs and correlated with patients’ survival. Results: Tumors showing SMA^high-, Periostin^highSMA^high-, or Periostin^highSMA^lowPDGFR^lowFAP^high-positive CAFs demonstrated significantly worse survival. Additionally, a high expression of SMA in PDAC tissue samples was shown to be an independent risk factor for worse survival. Conclusion: This project identified three subgroups of PDAC with different expression patterns of CAF markers which showed significantly worse survival. This could be the base for the further characterization of the fibroblast subgroups in PDAC and contribute to the development of new targeted therapy options against CAFs.

Purpose: Pancreatic ductal adenocarcinoma (PDAC) is currently one of the leading causes of cancer death worldwide. Therefore, building further subgroups as well as enabling individual patient therapy and diagnostics are needed. X-linked inhibitor of apoptosis protein (XIAP) is known to modulate apoptotic and inflammatory pathways. Its expression was found to correlate with patients’ survival in other tumor entities. This study aims to examine the role of XIAP in patients with PDAC in relation to the inflammatory microenvironment. Methods: The PANCALYZE multicenter study group included 257 patients with PDAC. Paraffin-embedded tumor samples were stained immunohistochemically for CD3, CD20, CD38, CD56, CD66b, CD117, and CD163 and XIAP. These stainings were further analyzed digitally with QuPath and survival analyses were done. Results: XIAP-positive patients with T-cell, respectively, neutrophil enriched tumors survived significantly longer compared to XIAP-negative patients (CD3: 37.6 vs. 24.6 months, p = 0.028; CD66b: 34.1 vs. 14.9 months, p = 0.027). Additionally, XIAP-positive patients showed better survival in the lymph node-negative population (48.4 vs. 24.2 months, p = 0.019). Regarding the total population, our findings did not show a correlation between XIAP expression and survival. In multivariate cox regression analyzes XIAP proves to be an independent factor for better survival in the identified subgroups (CD3: p = 0.043; CD66b: p = 0.012, N0: p = 0.040). Conclusion: We found XIAP-positive subgroups with significantly better survival in patients with PDAC in T-cell-rich, neutrophil-rich, or lymph node-negative cohorts. This could lead to further individualized cancer treatment with less aggressive therapy protocols for XIAP-positive tumors or more intensive follow-up for XIAP-negative tumors.

In pancreatic ductal adenocarcinoma (PDAC), the tumor stroma constitutes most of the cell mass and contributes to therapy resistance and progression. Here we show a hitherto unknown metabolic cooperation between pancreatic stellate cells (PSCs) and tumor cells through Interleukin 17B/Interleukin 17B receptor (IL-17B/IL-17RB) signaling. Tumor-derived IL-17B carrying extracellular vesicles (EVs) activated stromal PSCs and induced the expression of IL-17RB. PSCs increased oxidative phosphorylation while reducing mitochondrial turnover. PSCs activated tumor cells in a feedback loop. Tumor cells subsequently increased oxidative phosphorylation and decreased glycolysis partially via IL-6. In vivo, IL-17RB overexpression in PSCs accelerated tumor growth in a co-injection xenograft mouse model. Our results demonstrate a tumor-to-stroma feedback loop increasing tumor metabolism to accelerate tumor growth under optimal nutritional conditions.

Pancreatic cancer features elaborate mechanisms of immune evasion. The potential of new immune molecules was explored to restore the antitumor immune response. If these immune molecules are associated with poor survival, specific drugs could take effect. Here, we analyze the expression of VISTA, LAG3, IDO, and TIM3 on tumor-infiltrating lymphocytes (TILs) and its impact on patient survival. We analyzed 153 pancreatic cancer patients from the prospectively managed database of the multicentered PANCALYZE study. Immunohistochemistry on a tissue microarray assessed VISTA, LAG3, IDO, and TIM3 expression of TILs from the patients undergoing primary resection. Complementarily, we analyzed publicly available transcriptomic data (n = 903). Successful completion of chemotherapy, and lymph node status were independent predictors of survival in the multivariate analysis of the clinicopathologic parameters. Fifteen tumors were exclusively VISTA-positive, thirteen tumors expressed VISTA together with TIM3, and ten tumors expressed VISTA together with IDO. Patients featuring tumors with high numbers of IDO-positive TILs had better patient survival (p = 0.037). VISTA, LAG3, and TIM3 expression did not correlate with survival. The analysis of publicly available data did not show survival differences. Tumors rarely co-express more than two immune molecules at the same time, and VISTA is most frequently co-expressed. Although IDO generally inhibits T-cell proliferation, a high expression of IDO was associated with improved survival. We expect immune checkpoint inhibitors against VISTA, LAG3, and TIM3 to be inefficient in a clinical application.

Objectives The interferon-induced protein with multiple tetratricopeptide repeats 3 (IFIT3) seems to be associated with the prognosis in pancreatic cancer. Here we clarify whether the heterogeneity of IFIT3 expression affects previous IFIT3 analysis. Methods This retrospective study analyzes pancreatic cancer tissue samples retrieved by surgery from 2 independent patient cohorts. Patients underwent either primary surgery (n = 72) or received neoadjuvant chemotherapy (n = 12). Immunohistochemistry assessed IFIT3 expression and its heterogeneity. Complementarily, we analyzed publicly available transcriptomic data (n = 903). Results Of the primarily resected tumors, 16.4% were heterogeneous. Patients with IFIT3-negative tumors did not survive longer compared with patients with IFIT3-positive tumors. An analysis of publicly available data confirmed this result. Patients developing lung metastases had the best prognosis (4.8 years) with significantly lower IFIT3 expression compared with liver metastasis (P = 0.0117). Patients receiving neoadjuvant therapy who are IFIT3 negative had a longer disease-free survival (1.2 vs 0.3 years, P = 0.0081). Conclusions Low IFIT3 expression is not associated with longer survival. Divergent results from tissue microarray analyses could be explained with tumor heterogeneity. As a single biomarker, IFIT3 is not suitable for predicting disease prognosis. Recurrence of lung metastases and response to neoadjuvant chemotherapy are associated with low IFIT3 expression.

(1) Background: IL-17A accelerates pancreatic intraepithelial neoplasia (PanIN) progression. In this study, we examined whether IL-17A/IL-17RA promotes pancreatic ductal adenocarcinoma (PDAC) aggressiveness in terms of survival and cancer stem cell modulation. (2) Methods: In vitro, the wound-healing assay, the sphere formation assay, and flow cytometry were applied to assess cancer stem cell features. In vivo, pancreatic tumors were induced in C57BL/6 mice using electroporation with oncogenic plasmids (P53-/- R172H; KrasG12V). Anti-IL-17 antibodies were administered as immunotherapy. We analyzed IL-17A/IL-17RA related survival using publicly available transcriptomic data (n = 903). (3) Results: IL-17A/IL-17RA expression was not related to survival in PDAC patients. IL-17A neither induces stem cell markers nor increases sphere formation and cell motility in vitro. Blocking the IL-17A/IL-17RA axis in a murine pancreatic cancer model did not improve the survival of mice, but reduced the tumor burden slightly. (4) Conclusions: IL-17A does not promote stem cell expansion in PDAC cell lines. Blocking IL-17A/IL-17RA signaling does not interfere with pancreatic cancer development and progression and may not be considered as a promising monotherapy for PDAC.