SURVIVAL PREDICTION FOR BRAIN TUMOR PATIENTS USING GENE EXPRESSION DATA

Brain tumor is one of the most aggressive types of cancer in humans, with an estimated median survival time of 12 months and only 4% of the patients surviving more than 5 years after disease diagnosis. Until recently, brain tumor prognosis has been based only on clinical information such as tumor grade and patient age, but there are reports indicating that molecular profiling of gliomas can reveal subgroups of patients with distinct survival rates. We hypothesize that coupling molecular profiling of brain tumors with clinical information might improve predictions of patient survival time and, consequently, better guide future treatment decisions. In order to evaluate this hypothesis, the general goal of this research is to build models for survival prediction of glioma patients using DNA molecular profiles (U133 Affymetrix gene expression microarrays) along with clinical information. First, a predictive Random Forest model is built for binary outcomes (i.e. short vs. long-term survival) and a small subset of genes whose expression values can be used to predict survival time is selected. Following, a new statistical methodology is developed for predicting time-to-death outcomes using Bayesian ensemble trees. Due to a large heterogeneity observed within prognostic classes obtained by the Random Forest model, prediction can be improved by relating time-to-death with gene expression profile directly. We propose a Bayesian ensemble model for survival prediction which is appropriate for high-dimensional data such as gene expression data. Our approach is based on the ensemble "sum-of-trees" model which is flexible to incorporate additive and interaction effects between genes. We specify a fully Bayesian hierarchical approach and illustrate our methodology for the CPH, Weibull, and AFT survival models. We overcome the lack of conjugacy using a latent variable formulation to model the covariate effects which decreases computation time for model fitting. Also, our proposed models provides a model-free way to select important predictive prognostic markers based on controlling false discovery rates. We compare the performance of our methods with baseline reference survival methods and apply our methodology to an unpublished data set of brain tumor survival times and gene expression data, selecting genes potentially related to the development of the disease under study. A closing discussion compares results obtained by Random Forest and Bayesian ensemble methods under the biological/clinical perspectives and highlights the statistical advantages and disadvantages of the new methodology in the context of DNA microarray data analysis.

Comparison of OG1RF and an isogenic fsrB deletion mutant by transcriptional analysis: the Fsr system of Enterococcus faecalis is more than the activator of gelatinase and serine protease.

The FsrABC system of Enterococcus faecalis controls the expression of gelatinase and a serine protease via a quorum-sensing mechanism, and recent studies suggest that the Fsr system may also regulate other genes important for virulence. To investigate the possibility that Fsr influences the expression of additional genes, we used transcriptional profiling, with microarrays based on the E. faecalis strain V583 sequence, to compare the E. faecalis strain OG1RF with its isogenic mutant, TX5266, an fsrB deletion mutant. We found that the presence of an intact fsrB influences expression of numerous genes throughout the growth phases tested, namely, late log to early stationary phase. In addition, the Fsr regulon is independent of the activity of the proteases, GelE and SprE, whose expression was confirmed to be activated at all three time points tested. While expression of some genes (i.e., ef1097 and ef0750 to -757, encoding hypothetical proteins) was activated in late log phase in OG1RF versus the fsrB deletion mutant, expression of ef1617 to -1634 (eut-pdu orthologues) was highly repressed by the presence of an intact Fsr at entry into stationary phase. This is the first time that Fsr has been characterized as a negative regulator. The newly recognized Fsr-regulated targets include other factors, besides gelatinase, described as important for biofilms (BopD), and genes predicted to encode the surface proteins EF0750 to -0757 and EF1097, along with proteins implicated in several metabolic pathways, indicating that the FsrABC system may be an important regulator in strain OG1RF, with both positive and negative effects.

LDOC1, A NOVEL BIOMARKER OF PROGNOSIS IN CHRONIC LYMPHOCYTIC LEUKEMIA

In chronic lymphocytic leukemia (CLL), one of the best predictors of outcome is the somatic mutation status of the immunoglobulin heavy chain variable region (IGHV) genes. Patients whose CLL cells have unmutated IGHV genes have a median survival of 8 years; those with mutated IGHV genes have a median survival of 25 years. To identify new prognostic biomarkers and molecular targets for therapy in untreated CLL patients, we reanalyzed the raw data from four published gene expression profiling microarray studies. Of 88 candidate biomarkers associated with IGHV somatic mutation status, we identified LDOC1 (Leucine Zipper, Down-regulated in Cancer 1), as one of the most significantly differentially expressed genes that distinguished mutated from unmutated CLL cases. LDOC1 is a putative transcription factor of unknown function in B-cell development and CLL pathophysiology. Using a highly sensitive quantitative RT-PCR (QRT-PCR) assay, we confirmed that LDOC1 mRNA was dramatically down-regulated in mutated compared to unmutated CLL cases. Expression of LDOC1 mRNA was also vii strongly associated with other markers of poor prognosis, including ZAP70 protein and cytogenetic abnormalities of poor prognosis (deletions of chromosomes 6q21, 11q23, and 17p13.1, and trisomy 12). CLL cases positive for LDOC1 mRNA had significantly shorter overall survival than negative cases. Moreover, in a multivariate model, LDOC1 mRNA expression predicted overall survival better than IGHV mutation status or ZAP70 protein, among the best markers of prognosis in CLL. We also discovered LDOC1S, a new LDOC1 splice variant. Using isoform-specific QRT-PCR assays that we developed, we found that both isoforms were expressed in normal B cells (naïve > memory), unmutated CLL cells, and in B-cell non-Hodgkin lymphomas with unmutated IGHV genes. To investigate pathways in which LDOC1 is involved, we knocked down LDOC1 in HeLa cells and performed global gene expression profiling. GFI1 (Growth Factor-Independent 1) emerged as a significantly up-regulated gene in both HeLa cells and CLL cells that expressed high levels of LDOC1. GFI1 oncoprotein is implicated in hematopoietic stem cell maintenance, lymphocyte development, and lymphomagenesis. Our findings indicate that LDOC1 mRNA is an excellent biomarker of overall survival in CLL, and may contribute to B-cell differentiation and malignant transformation.

Emdogain-regulated gene expression in palatal fibroblasts requires TGF-βRI kinase signaling.

Genome-wide microarrays have suggested that Emdogain regulates TGF-β target genes in gingival and palatal fibroblasts. However, definitive support for this contention and the extent to which TGF-β signaling contributes to the effects of Emdogain has remained elusive. We therefore studied the role of the TGF-β receptor I (TGF-βRI) kinase to mediate the effect of Emdogain on palatal fibroblasts. Palatal fibroblasts were exposed to Emdogain with and without the inhibitor for TGF-βRI kinase, SB431542. Emdogain caused 39 coding genes to be differentially expressed in palatal fibroblasts by microarray analysis (p<0.05; >10-fold). Importantly, in the presence of the TGF-βRI kinase inhibitor SB431542, Emdogain failed to cause any significant changes in gene expression. Consistent with this mechanism, three independent TGF-βRI kinase inhibitors and a TGF-β neutralizing antibody abrogated the increased expression of IL-11, a selected Emdogain target gene. The MAPK inhibitors SB203580 and U0126 lowered the impact of Emdogain on IL-11 expression. The data support that TGF-βRI kinase activity is necessary to mediate the effects of Emdogain on gene expression in vitro.

A next-generation tissue microarray (ngTMA) protocol for biomarker studies

Biomarker research relies on tissue microarrays (TMA). TMAs are produced by repeated transfer of small tissue cores from a 'donor' block into a 'recipient' block and then used for a variety of biomarker applications. The construction of conventional TMAs is labor intensive, imprecise, and time-consuming. Here, a protocol using next-generation Tissue Microarrays (ngTMA) is outlined. ngTMA is based on TMA planning and design, digital pathology, and automated tissue microarraying. The protocol is illustrated using an example of 134 metastatic colorectal cancer patients. Histological, statistical and logistical aspects are considered, such as the tissue type, specific histological regions, and cell types for inclusion in the TMA, the number of tissue spots, sample size, statistical analysis, and number of TMA copies. Histological slides for each patient are scanned and uploaded onto a web-based digital platform. There, they are viewed and annotated (marked) using a 0.6-2.0 mm diameter tool, multiple times using various colors to distinguish tissue areas. Donor blocks and 12 'recipient' blocks are loaded into the instrument. Digital slides are retrieved and matched to donor block images. Repeated arraying of annotated regions is automatically performed resulting in an ngTMA. In this example, six ngTMAs are planned containing six different tissue types/histological zones. Two copies of the ngTMAs are desired. Three to four slides for each patient are scanned; 3 scan runs are necessary and performed overnight. All slides are annotated; different colors are used to represent the different tissues/zones, namely tumor center, invasion front, tumor/stroma, lymph node metastases, liver metastases, and normal tissue. 17 annotations/case are made; time for annotation is 2-3 min/case. 12 ngTMAs are produced containing 4,556 spots. Arraying time is 15-20 hr. Due to its precision, flexibility and speed, ngTMA is a powerful tool to further improve the quality of TMAs used in clinical and translational research.

Novel prognostic markers revealed by a proteomic approach separating benign from malignant insulinomas.

The prognosis of pancreatic neuroendocrine tumors is related to size, histology and proliferation rate. However, this stratification needs to be refined further. We conducted a proteome study on insulinomas, a well-defined pancreatic neuroendocrine tumor entity, in order to identify proteins that can be used as biomarkers for malignancy. Based on a long follow-up, insulinomas were divided into those with metastases (malignant) and those without (benign). Microdissected cells from six benign and six malignant insulinomas were subjected to a procedure combining fluorescence dye saturation labeling with high-resolution two-dimensional gel electrophoresis. Differentially expressed proteins were identified using nano liquid chromatography-electrospray ionization/multi-stage mass spectrometry and validated by immunohistochemistry on tissue microarrays containing 62 insulinomas. Sixteen differentially regulated proteins were identified among 3000 protein spots. Immunohistochemical validation revealed that aldehyde dehydrogenase 1A1 and voltage-dependent anion-selective channel protein 1 showed significantly stronger expression in malignant insulinomas than in benign insulinomas, whereas tumor protein D52 (TPD52) binding protein was expressed less strongly in malignant insulinomas than in benign insulinomas. Using multivariate analysis, low TPD52 expression was identified as a strong independent prognostic factor for both recurrence-free and overall disease-related survival.

MicroRNA profile of poorly differentiated thyroid carcinomas: new diagnostic and prognostic insights.

The diagnosis of conventional and oncocytic poorly differentiated (oPD) thyroid carcinomas is difficult. The aim of this study is to characterise their largely unknown miRNA expression profile and to compare it with well-differentiated thyroid tumours, as well as to identify miRNAs which could potentially serve as diagnostic and prognostic markers. A total of 14 poorly differentiated (PD), 13 oPD, 72 well-differentiated thyroid carcinomas and eight normal thyroid specimens were studied for the expression of 768 miRNAs using PCR-Microarrays. MiRNA expression was different between PD and oPD thyroid carcinomas, demonstrating individual clusters on the clustering analysis. Both tumour types showed upregulation of miR-125a-5p, -15a-3p, -182, -183-3p, -222, -222-5p, and downregulation of miR-130b, -139-5p, -150, -193a-5p, -219-5p, -23b, -451, -455-3p and of miR-886-3p as compared with normal thyroid tissue. In addition, the oPD thyroid carcinomas demonstrated upregulation of miR-221 and miR-885-5p. The difference in expression was also observed between miRNA expression in PD and well-differentiated tumours. The CHAID algorithm allowed the separation of PD from well-differentiated thyroid carcinomas with 73-79% accuracy using miR-23b and miR-150 as a separator. Kaplan-Meier and multivariate analysis showed a significant association with tumour relapses (for miR-23b) and with tumour-specific death (for miR-150) in PD and oPD thyroid carcinomas. MiRNA expression is different in conventional and oPD thyroid carcinomas in comparison with well-differentiated thyroid cancers and can be used for discrimination between these tumour types. The newly identified deregulated miRNAs (miR-150, miR-23b) bear the potential to be used in a clinical setting, delivering prognostic and diagnostic informations.

A novel universal DNA labeling and amplification system for rapid microarray-based detection of 117 antibiotic resistance genes in Gram-positive bacteria.

A rapid and simple DNA labeling system has been developed for disposable microarrays and has been validated for the detection of 117 antibiotic resistance genes abundant in Gram-positive bacteria. The DNA was fragmented and amplified using phi-29 polymerase and random primers with linkers. Labeling and further amplification were then performed by classic PCR amplification using biotinylated primers specific for the linkers. The microarray developed by Perreten et al. (Perreten, V., Vorlet-Fawer, L., Slickers, P., Ehricht, R., Kuhnert, P., Frey, J., 2005. Microarray-based detection of 90 antibiotic resistance genes of gram-positive bacteria. J.Clin.Microbiol. 43, 2291-2302.) was improved by additional oligonucleotides. A total of 244 oligonucleotides (26 to 37 nucleotide length and with similar melting temperatures) were spotted on the microarray, including genes conferring resistance to clinically important antibiotic classes like β-lactams, macrolides, aminoglycosides, glycopeptides and tetracyclines. Each antibiotic resistance gene is represented by at least 2 oligonucleotides designed from consensus sequences of gene families. The specificity of the oligonucleotides and the quality of the amplification and labeling were verified by analysis of a collection of 65 strains belonging to 24 species. Association between genotype and phenotype was verified for 6 antibiotics using 77 Staphylococcus strains belonging to different species and revealed 95% test specificity and a 93% predictive value of a positive test. The DNA labeling and amplification is independent of the species and of the target genes and could be used for different types of microarrays. This system has also the advantage to detect several genes within one bacterium at once, like in Staphylococcus aureus strain BM3318, in which up to 15 genes were detected. This new microarray-based detection system offers a large potential for applications in clinical diagnostic, basic research, food safety and surveillance programs for antimicrobial resistance.

Circulating microRNA profiling in patients with advanced non-squamous NSCLC receiving bevacizumab/erlotinib followed by platinum-based chemotherapy at progression (SAKK 19/05).

OBJECTIVES Molecular subclassification of non small-cell lung cancer (NSCLC) is essential to improve clinical outcome. This study assessed the prognostic and predictive value of circulating micro-RNA (miRNA) in patients with non-squamous NSCLC enrolled in the phase II SAKK (Swiss Group for Clinical Cancer Research) trial 19/05, receiving uniform treatment with first-line bevacizumab and erlotinib followed by platinum-based chemotherapy at progression. MATERIALS AND METHODS Fifty patients with baseline and 24 h blood samples were included from SAKK 19/05. The primary study endpoint was to identify prognostic (overall survival, OS) miRNA's. Patient samples were analyzed with Agilent human miRNA 8x60K microarrays, each glass slide formatted with eight high-definition 60K arrays. Each array contained 40 probes targeting each of the 1347 miRNA. Data preprocessing included quantile normalization using robust multi-array average (RMA) algorithm. Prognostic and predictive miRNA expression profiles were identified by Spearman's rank correlation test (percentage tumor shrinkage) or log-rank testing (for time-to-event endpoints). RESULTS Data preprocessing kept 49 patients and 424 miRNA for further analysis. Ten miRNA's were significantly associated with OS, with hsa-miR-29a being the strongest prognostic marker (HR=6.44, 95%-CI 2.39-17.33). Patients with high has-miR-29a expression had a significantly lower survival at 10 months compared to patients with a low expression (54% versus 83%). Six out of the 10 miRNA's (hsa-miRN-29a, hsa-miR-542-5p, hsa-miR-502-3p, hsa-miR-376a, hsa-miR-500a, hsa-miR-424) were insensitive to perturbations according to jackknife cross-validation on their HR for OS. The respective principal component analysis (PCA) defined a meta-miRNA signature including the same 6 miRNA's, resulting in a HR of 0.66 (95%-CI 0.53-0.82). CONCLUSION Cell-free circulating miRNA-profiling successfully identified a highly prognostic 6-gene signature in patients with advanced non-squamous NSCLC. Circulating miRNA profiling should further be validated in external cohorts for the selection and monitoring of systemic treatment in patients with advanced NSCLC.

Proper paraffin slide storage is crucial for translational research projects involving immunohistochemistry stains.

The use of paraffin slides and tissue microarrays (TMA) is indispensable for translational research. However, storage of paraffin slides over time has a substantial detrimental effect on the quality and reliability of immunohistochemistry stains. Particularly affected by this issue may be any collaborative efforts where paraffin slides or TMAs are shipped to central laboratories and then 'biobanked' for some time until use. This article summarizes some of the key issues affecting loss of antigenicity on paraffin slides and some simple storage solutions to help maintain high quality immunohistochemistry results when paraffin slides must be stored for a certain time prior to use.

HLA class II antigen expression in colorectal carcinoma tumors as a favorable prognostic marker.

The goal of this study was to determine the frequency of HLA class II antigen expression in colorectal carcinoma (CRC) tumors, its association with the clinical course of the disease, and the underlying mechanism(s). Two tissue microarrays constructed with 220 and 778 CRC tumors were stained with HLA-DR, DQ, and DP antigen-specific monoclonal antibody LGII-612.14, using the immunoperoxidase staining technique. The immunohistochemical staining results were correlated with the clinical course of the disease. The functional role of HLA class II antigens expressed on CRC cells was analyzed by investigating their in vitro interactions with immune cells. HLA class II antigens were expressed in about 25% of the 220 and 21% of the 778 tumors analyzed with an overall frequency of 23%. HLA class II antigens were detected in 19% of colorectal adenomas. Importantly, the percentage of stained cells and the staining intensity were significantly lower than those detected in CRC tumors. However, HLA class II antigen staining was weakly detected only in 5.4% of 37 normal mucosa tissues. HLA class II antigen expression was associated with a favorable clinical course of the disease. In vitro stimulation with interferon gamma (IFNγ) induced HLA class II antigen expression on two of the four CRC cell lines tested. HLA class II antigen expression on CRC cells triggered interleukin-1β (IL-1β) production by resting monocytes. HLA class II antigen expression in CRC tumors is a favorable prognostic marker. This association may reflect stimulation of IL-1β production by monocytes.

Bcl-2 predicts response to neoadjuvant chemotherapy and is overexpressed in lymph node metastases of urothelial cancer of the bladder

PURPOSE To assess whether Bcl-2, an inhibitor of the apoptotic cascade, can predict response to neoadjuvant chemotherapy in patients with urothelial cancer of the bladder (UCB). METHODS Bcl-2 expression was analyzed in 2 different tissue microarrays (TMAs). One TMA was constructed of primary tumors and their corresponding lymph node (LN) metastases from 152 patients with chemotherapy-naive UCB treated by cystectomy and pelvic lymphadenectomy (chemotherapy-naive TMA cohort). The other TMA was constructed of tumor samples obtained from 55 patients with UCB before neoadjuvant chemotherapy (transurethral resection of the bladder cancer) and after cystectomy with pelvic lymphadenectomy (residual primary tumor [ypT+], n = 38); residual LN metastases [ypN+], n = 24) (prechemotherapy/postchemotherapy TMA cohort). Bcl-2 overexpression was defined as 10% or more cancer cells showing cytoplasmic immunoreactivity. RESULTS In both TMA cohorts, Bcl-2 overexpression was significantly (P<0.05) more frequent in LN metastases than in primary tumors (chemotherapy-naive TMA group: 18/148 [12%] in primary tumors vs. 39/143 [27%] in metastases; postchemotherapy TMA: ypT+7/35 [20%] vs. ypN+11/19 [58%]). In the neoadjuvant setting, patients with Bcl-2 overexpression in transurethral resection of the bladder cancer specimens showed significantly (P = 0.04) higher ypT stages and less regression in their cystectomy specimens than did the control group, and only one-eighth (13%) had complete tumor regression (ypT0 ypN0). In survival analyses, only histopathological parameters added significant prognostic information. CONCLUSIONS Bcl-2 overexpression in chemotherapy-naive primary bladder cancer is related to poor chemotherapy response and might help to select likely nonresponders.

Image analysis of immunohistochemistry is superior to visual scoring as shown for patient outcome of esophageal adenocarcinoma.

Quantification of protein expression based on immunohistochemistry (IHC) is an important step in clinical diagnoses and translational tissue-based research. Manual scoring systems are used in order to evaluate protein expression based on staining intensities and distribution patterns. However, visual scoring remains an inherently subjective approach. The aim of our study was to explore whether digital image analysis proves to be an alternative or even superior tool to quantify expression of membrane-bound proteins. We analyzed five membrane-binding biomarkers (HER2, EGFR, pEGFR, β-catenin, and E-cadherin) and performed IHC on tumor tissue microarrays from 153 esophageal adenocarcinomas patients from a single center study. The tissue cores were scored visually applying an established routine scoring system as well as by using digital image analysis obtaining a continuous spectrum of average staining intensity. Subsequently, we compared both assessments by survival analysis as an end point. There were no significant correlations with patient survival using visual scoring of β-catenin, E-cadherin, pEGFR, or HER2. In contrast, the results for digital image analysis approach indicated that there were significant associations with disease-free survival for β-catenin, E-cadherin, pEGFR, and HER2 (P = 0.0125, P = 0.0014, P = 0.0299, and P = 0.0096, respectively). For EGFR, there was a greater association with patient survival when digital image analysis was used compared to when visual scoring was (visual: P = 0.0045, image analysis: P < 0.0001). The results of this study indicated that digital image analysis was superior to visual scoring. Digital image analysis is more sensitive and, therefore, better able to detect biological differences within the tissues with greater accuracy. This increased sensitivity improves the quality of quantification.

Normal platelet activation profile in patients with peripheral arterial disease on aspirin.

BACKGROUND Peripheral arterial disease (PAD) is a progressive vascular disease associated with a high risk of cardiovascular morbidity and death. Antithrombotic prevention is usually applied by prescribing the antiplatelet agent aspirin. However, in patients with PAD aspirin fails to provide protection against myocardial infarction and death, only reducing the risk of ischemic stroke. Platelets may play a role in disease development, but this has not been tested by proper mechanistic studies. In the present study, we performed a systematic evaluation of platelet reactivity in whole blood from patients with PAD using two high-throughput assays, i.e. multi-agonist testing of platelet activation by flow cytometry and multi-parameter testing of thrombus formation on spotted microarrays. METHODS Blood was obtained from 40 patients (38 on aspirin) with PAD in majority class IIa/IIb and from 40 age-matched control subjects. Whole-blood flow cytometry and multiparameter thrombus formation under high-shear flow conditions were determined using recently developed and validated assays. RESULTS Flow cytometry of whole blood samples from aspirin-treated patients demonstrated unchanged high platelet responsiveness towards ADP, slightly elevated responsiveness after glycoprotein VI stimulation, and decreased responsiveness after PAR1 thrombin receptor stimulation, compared to the control subjects. Most parameters of thrombus formation under flow were similarly high for the patient and control groups. However, in vitro aspirin treatment caused a marked reduction in thrombus formation, especially on collagen surfaces. When compared per subject, markers of ADP- and collagen-induced integrin activation (flow cytometry) strongly correlated with parameters of collagen-dependent thrombus formation under flow, indicative of a common, subject-dependent regulation of both processes. CONCLUSION Despite of the use of aspirin, most platelet activation properties were in the normal range in whole-blood from class II PAD patients. These data underline the need for more effective antithrombotic pharmacoprotection in PAD.

Accumulation of FOXP3+T-cells in the tumor microenvironment is associated with an epithelial-mesenchymal-transition-type tumor budding phenotype and is an independent prognostic factor in surgically resected pancreatic ductal adenocarcinoma.

Here we explore the role of the interplay between host immune response and epithelial-mesenchymal-transition (EMT)-Type tumor-budding on the outcome of pancreatic adenocarcinoma (PDAC).CD4+, CD8+, and FOXP3+T-cells as well as iNOS+ (M1) and CD163+- macrophages (M2) were assessed on multipunch tissue-microarrays containing 120 well-characterized PDACs, precursor lesions (PanINs) and corresponding normal tissue. Counts were normalized for the percentage of tumor/spot and associated with the clinico-pathological features, including peritumoral (PTB) and intratumoral (ITB) EMT-Type tumor-budding and outcome.Increased FOXP3+T-cell-counts and CD163-macrophages and decreased CD8+T-cell-counts were observed in PDACs compared with normal tissues and PanINs (p < 0.0001). Increased peritumoral FOXP3+T-cell-counts correlated significantly with venous invasion, distant metastasis, R1-status, high-grade ITB, PTB and independently with reduced survival. Increased intratumoral FOXP3+T-cells correlated with lymphatic invasion, N1-stage, PTB and marginally with adverse outcome. High peritumoral CD163-counts correlated with venous invasion, PTB and ITB. High intratumoral CD163-counts correlated with higher T-stage and PTB.PDAC-microenvironment displays a tumor-favoring immune-cell composition especially in the immediate environment of the tumor-buds that promotes further growth and indicates a close interaction of the immune response with the EMT-process. Increased peritumoral FOXP3+T-cell density is identified as an independent adverse prognostic factor in PDAC. Patients with phenotypically aggressive PDACs may profit from targeted immunotherapy against FOXP3.