Marker selection strategies for circulating tumor DNA guided by phylogenetic inference

Motivation

Blood-based profiling of tumor DNA ("liquid biopsy") has offered great prospects for non-invasive early cancer diagnosis, treatment monitoring, and clinical guidance, but require further advances in computational methods to become a robust quantitative assay of tumor clonal evolution. We propose new methods to better characterize tumor clonal dynamics from circulating tumor DNA (ctDNA), through application to two specific questions: 1) How to apply longitudinal ctDNA data to refine phylogeny models of clonal evolution, and 2) how to quantify changes in clonal frequencies that may be indicative of treatment response or tumor progression. We pose these questions through a probabilistic framework for optimally identifying maximum likelihood markers and applying them to characterizing clonal evolution.

Results

We first estimate a distribution over plausible clonal lineage models, using bootstrap samples over pre-treatment tissue-based sequence data. We then refine these lineage models and the clonal frequencies they imply over successive longitudinal samples. We use the resulting framework for modeling and refining tree distributions to pose a set of optimization problems to select ctDNA markers to maximize measures of utility capturing ability to solve the two questions of reducing uncertain in phylogeny models or quantifying clonal frequencies given the models. We tested our methods on synthetic data and showed them to be effective at refining distributions of tree models and clonal frequencies so as to minimize measures of tree distance relative to the ground truth. Application of the tree refinement methods to real tumor data further demonstrated their effectiveness in refining a clonal lineage model and assessing its clonal frequencies. The work shows the power of computational methods to improve marker selection, clonal lineage reconstruction, and clonal dynamics profiling for more precise and quantitative assays of tumor progression.

Availability

https://github.com/CMUSchwartzLab/Mase-phi.git.

Contact

russells@andrew.cmu.edu.

Modeling the Effect of Spatial Structure on Solid Tumor Evolution and Circulating Tumor DNA Composition

Circulating tumor DNA (ctDNA) monitoring, while sufficiently advanced to reflect tumor evolution in real time and inform cancer diagnosis, treatment, and prognosis, mainly relies on DNA that originates from cell death via apoptosis or necrosis. In solid tumors, chemotherapy and immune infiltration can induce spatially variable rates of cell death, with the potential to bias and distort the clonal composition of ctDNA. Using a stochastic evolutionary model of boundary-driven growth, we study how elevated cell death on the edge of a tumor can simultaneously impact driver mutation accumulation and the representation of tumor clones and mutation detectability in ctDNA. We describe conditions in which invasive clones are over-represented in ctDNA, clonal diversity can appear elevated in the blood, and spatial bias in shedding can inflate subclonal variant allele frequencies (VAFs). Additionally, we find that tumors that are mostly quiescent can display similar biases but are far less detectable, and the extent of perceptible spatial bias strongly depends on sequence detection limits. Overall, we show that spatially structured shedding might cause liquid biopsies to provide highly biased profiles of tumor state. While this may enable more sensitive detection of expanding clones, it could also increase the risk of targeting a subclonal variant for treatment. Our results indicate that the effects and clinical consequences of spatially variable cell death on ctDNA composition present an important area for future work.

An 11-gene expression signature related to tumorigenesis and immunosuppression in primary cutaneous melanoma predicts sentinel lymph node metastatic status

e21579

Background: A biomarker derived from the primary melanoma tumor to predict regional sentinel lymph node (SLN) metastatic status can be valuable in guiding the decision making in planning the SLN surgical procedure for candidate patients. Methods: Gene expression profiling was performed on primary cutaneous melanoma tumor biopsies of 49 (24 known SLN+, 25 SLN-) patients (T3a/b, T4a/b) who underwent SLN for staging using transcriptome profiling analysis. U133A 2.0 Affymetrix gene chips were used. Significance Analysis of Microarrays (SAM) was used to test the association between gene expression level of the primary tumor and SLN status. Genes with fold change > 1.5 and q value < 0.05 were considered differentially expressed. Pathway analysis was performed using Ingenuity Pathway Analysis. Benjamini and Hochberg method was used to adjust for multiple testing in pathway analysis. All statistical analyses were implemented in R. Results: A total of 49 patients with primary cutaneous melanoma were studied, of which, 24 were diagnosed as SLN positive and 25 as SLN negative by routine H&E and immunohistochemistry. Using SLN metastatic status as the outcome, a univariate logistic regression model was fitted with individual probe sets. A total of 251 probe sets were filtered and 11 probe sets were considered as differentially expressed (DE) between SLN-ve and SLN+ve groups, with the selection criterion set at Benjamini- Hochberg method -adjusted p -value below 0.05 and the absolute log fold change above 0.5. As each of the 11 probe sets was matched to a unique known gene, an 11 gene signature was derived and among them, the expression level of 7 genes was significantly reduced in the SLN+ve group compared to the SLN-ve group, while 4 genes were overexpressed in the SLN+ve group. Integrative and interactive heatmaps were produced from hierarchical cluster analysis to show the 11 differentially expressed genes. Selected genes were found to be uniformly related to tumorigenesis, malignant progression, DNA repair, cell cycle regulation, chemoresistance, immunosuppression and/or involvement in multiple cancer-related pathways. Several of the selected genes were previously shown to be prognostic in various malignancies. Conclusions: We present a unique 11-gene expression signature derived from primary melanoma tumor biopsy samples and related to tumorigenesis and immunosuppression that may be useful for the prognostic stratification of melanoma patients. This may allow the optimal selection of patients for SLN biopsy.

88 Evidence of enhanced immune activation within the tumor microenvironment and the circulation of female patients with high-risk melanoma compared to males

Background

Sex differences in tumor immunity and response to immunotherapy were shown in murine models and descriptive analyses from recent clinical trials. We recently reported that female gender is a favorable prognostic marker for survival benefit following ipilimumab and high dose interferon-alfa (HDI) adjuvant therapy of high-risk melanoma in the ECOG-ACRIN E1609 trial (N=1670). Therefore, we investigated differences in candidate immune biomarkers in the circulation and tumor microenvironment (TME) of female and male patients.

Methods

Gene expression profiling (GEP) was performed on the tumor biopsies of 718 (454 male, 264 female) patients. The primary endpoint was mRNA expression profiling using U133A 2.0 Affymetrix gene chips. Raw microarray data sets were normalized by using the Robust Multi-array Average (RMA) method using Affymetrix Power Tools (APT) as previously published. Multiple probe sets representing the same genes were collapsed by using the probe with maximum gene expression. Gene set enrichment analysis (GSEA) was performed by comparing the female and male tumor samples, and gene sets with FDR q-value <0.1 were deemed as significant. Similarly, peripheral blood (serum and PBMC) samples were tested for soluble (Luminex) and cellular (multicolor flow cytometry) prognostic biomarkers in a subset of patients (N=321; 109 female and 212 male). All patients provided an IRB-approved written informed consent.

Results

Among the subset of patients tested for circulating biomarkers, females were significantly younger than males (P=0.03). Testing PBMCs, the percentages of CD3+ T cells (P=0.04) and CD3+CD4+ helper T cells (P=0.0005) were significantly higher in female patients compared to males. Also, there were trends toward higher levels of proinflammatory cytokines IL1beta (P=0.07) and IL6 (P=0.06) in females. On the other hand, males had significantly higher percentages of monocytes (P=0.03). Further, there were trends toward higher percentages of CD3+/CD4+/CD25hi+/Foxp3+ (P=0.1) and CD3+/CD4+/CD25+/CD127low+ (P=0.1) T-reg in male patients compared to females. Among the cohort of patients (N=718) with tumor GEP data, females were significantly younger than males (P=0.0009). GEP identified pathways and genes related to immune cell infiltration and activation that were significantly enriched in the tumors of females compared to males (table 1).

Abstract 88 Table 1

Immune pathways significantly enriched in tumors of females

Conclusions

Female gender was associated with adjuvant immunotherapeutic benefits and female patients were more likely to have evidence of immune activation within the TME and the circulation, supporting a potentially important role for female related factors in the immune response against melanoma, and these require further investigation.

Acknowledgements

We are grateful to the patients and family members who participated in the E1609 trial and the E1609 trial investigators. This study was conducted by the ECOG-ACRIN Cancer Research Group (Peter J. O’Dwyer, MD and Mitchell D. Schnall, MD, PhD, Group Co-Chairs) and supported by the National Cancer Institute of the National Institutes of Health under the following award numbers: U10CA180794, U10CA180820, U10CA180888, UG1CA233180, UG1CA233184. Biomarkers studies were supported under the following award number: P50CA12197310 (Tarhini and Kirkwood). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Trial Registration

NCT01274338

Ethics Approval

The E1609 study protocol was approved by the institutional review board of each participating institution and conducted in accordance with Good Clinical Practice guidelines as defined by the International Conference on Harmonization. All patients provided an IRB-approved written informed consent.

Consent

Not applicable.

87 Enhanced immunogenicity within the tumor microenvironment and the circulation of high-risk melanoma patients with unknown primary compared to those whose primary melanoma is known

Background

We recently reported data supporting the unknown primary status as a potentially distinct prognostic group among high-risk melanoma patients treated with ipilimumab and high dose interferon-alfa (HDI) in the ECOG-ACRIN E1609 trial (N=1670) with improved RFS and OS outcomes compared to known primary. Therefore, we investigated differences in candidate immune biomarkers in the circulation and tumor microenvironment (TME) of patients with unknown compared to those with known primary melanoma enrolled in this trial that tested adjuvant ipilimumab at 3 and 10 mg/kg versus HDI.

Methods

Gene expression profiling (GEP) was performed on the tumor biopsies of 718 (102 unknown, 616 known primary) melanoma patients. The primary endpoint was mRNA expression profiling using U133A 2.0 Affymetrix gene chips. Raw microarray data sets were normalized by using the Robust Multi-array Average (RMA) method using Affymetrix Power Tools (APT) as previously published. Multiple probe sets representing the same genes were collapsed by using the probe with maximum gene expression. Gene set enrichment analysis (GSEA) was performed by comparing the unknown and known primary tumor samples, and gene sets with FDR q-value <0.1 were deemed as significant. Similarly, peripheral blood (serum and PBMC) samples were tested for soluble (Luminex) and cellular (multicolor flow cytometry) immune biomarkers in a subset of patients (N=321; 66 unknown and 255 known primary). All patients provided an IRB-approved written informed consent.

Results

Unknown primary melanoma cases represented 12.8% of the total E1609 study population (N=1670) including 11.7% on the ipilimumab arms and 14.7% on the HDI arm. Stratifying by stage, relapse free survival (RFS) (P=0.001) and overall survival (OS) (P=0.009) were significantly better for patients with unknown primary tumor compared to known primary. Including only ipilimumab-treated patients, RFS (P=0.005) and OS (P=0.023) were significantly better in favor of the unknown primary status. Among the cohort of patients with tumor GEP data (N=718), GEP identified pathways and genes related to autoimmunity, inflammation, immune cell infiltration and immune activation that were significantly enriched in the unknown primary tumors compared to known primaries (table 1). Among the subset of patients tested for circulating biomarkers, patients with unknown primary melanoma had significantly higher circulating levels of IL-2R than those with known primary (P=0.04).

Abstract 87 Table 1

Immune pathways enriched in unknown primary melanoma

Conclusions

Unknown primary high-risk melanoma patients had significantly better prognosis and evidence of significantly enhanced immune activation within the TME and the circulation, supporting the designation of unknown primary melanoma as a distinct prognostic marker in patients with high-risk melanoma.

Acknowledgements

We are grateful to the patients and family members who participated in the E1609 trial and the E1609 trial investigators. This study was conducted by the ECOG-ACRIN Cancer Research Group (Peter J. O’Dwyer, MD and Mitchell D. Schnall, MD, PhD, Group Co-Chairs) and supported by the National Cancer Institute of the National Institutes of Health under the following award numbers: U10CA180794, U10CA180820, U10CA180888, UG1CA233180, UG1CA233184. Biomarkers studies were supported under the following award number: P50CA12197310 (Tarhini and Kirkwood). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Trial Registration

NCT01274338

Ethics Approval

The E1609 study protocol was approved by the institutional review board of each participating institution and conducted in accordance with Good Clinical Practice guidelines as defined by the International Conference on Harmonization. All patients provided an IRB-approved written informed consent.

Consent

Not applicable.

Absolute quantitative proteomics using the total protein approach to identify novel clinical immunohistochemical markers in renal neoplasms

BACKGROUND

Renal neoplasms encompass a variety of malignant and benign tumors, including many with shared characteristics. The diagnosis of these renal neoplasms remains challenging with currently available tools. In this work, we demonstrate the total protein approach (TPA) based on high-resolution mass spectrometry (MS) as a tool to improve the accuracy of renal neoplasm diagnosis.

METHODS

Frozen tissue biopsies of human renal tissues [clear cell renal cell carcinoma (n = 7), papillary renal cell carcinoma (n = 5), chromophobe renal cell carcinoma (n = 5), and renal oncocytoma (n = 5)] were collected for proteome analysis. Normal adjacent renal tissue (NAT, n = 5) was used as a control. Proteins were extracted and digested using trypsin, and the digested proteomes were analyzed by label-free high-resolution MS (nanoLC-ESI-HR-MS/MS). Quantitative analysis was performed by comparison between protein abundances of tumors and NAT specimens, and the label-free and standard-free TPA was used to obtain absolute protein concentrations.

RESULTS

A total of 205 differentially expressed proteins with the potential to distinguish the renal neoplasms were found. Of these proteins, a TPA-based panel of 24, including known and new biomarkers, was selected as the best candidates to differentiate the neoplasms. As proof of concept, the diagnostic potential of PLIN2, TUBB3, LAMP1, and HK1 was validated using semi-quantitative immunohistochemistry with a total of 128 samples assessed on tissue micro-arrays.

CONCLUSIONS

We demonstrate the utility of combining high-resolution MS and the TPA as potential new diagnostic tool in the pathology of renal neoplasms. A similar TPA approach may be implemented in any cancer study with solid biopsies.

Somatic mutations predict immune-related adverse events (irAEs) and survival following CTLA4 blockade in melanoma patients

e21584

Background: Predicting immune-related adverse events (irAEs) resulting from immune checkpoint inhibitors (ICIs) as well as the likelihood of clinical benefit is an area of great interest. irAEs have been associated with clinical outcome measures and we were interested in examining somatic mutations that may correlate with irAEs and subsequently with relapse free survival (RFS) and overall survival (OS). Methods: Our irAEs of interest for this analysis were rash and colitis based on prior data and these were tested for association with RFS and overall OS. Nonsynonymous somatic mutations were identified utilizing tumor and matched blood samples from 22 metastatic melanoma patients treated with neoadjuvant ipilimumab. Whole exome sequencing was performed using the SOLiD 5500 Next-Generation Sequencing platform. Somatic mutations were identified by subtraction of germline variants. Total mutational burden was assessed, and individual mutations were correlated with irAEs and with clinical outcome. Results: MUC16 was the most commonly mutated gene (82%). Of notable known oncogenic mutations, BRAF was the most common (50%) followed by NRAS (23%). Among known oncogenic signaling pathways, enrichment was most common in RTK-RAS (58%), followed by NOTCH and WNT. Development of colitis showed a strong trend in association with RFS. Development of rash indicated a similar trend. 20 different statistically significant mutations (p < 0.05) were identified among patients with colitis and these were also associated with OS and RFS. Notably, PTPRO mutation status (known to antagonize PDL1 expression) appeared protective against colitis and associated with worse RFS and OS. Similarly, ADGB mutation was significantly associated with developing rash and also associated with improved RFS and OS. Correcting for TMB in a multivariate model confirmed the associations. Conclusions: Somatic mutations were found to be associated with the irAEs of rash and colitis and with clinical outcome following ipilimumab treatment of melanoma. Our findings warrant further investigation and validation in independent cohorts.

Ultrasonic-assisted extraction and digestion of proteins from solid biopsies followed by peptide sequential extraction hyphenated to MALDI-based profiling holds the promise of distinguishing renal oncocytoma from chromophobe renal cell carcinoma

A novel analytical approach is proposed to discriminate between solid biopsies of chromophobe renal cell carcinoma (chRCC) and renal oncocytoma (RO). The method comprises the following steps: (i) ultrasonic extraction of proteins from solid biopsies, (ii) protein depletion with acetonitrile, (iii) ultrasonic assisted in-solution digestion using magnetic nanoparticle with immobilized trypsin, (iv) C18 tip-based preconcentration of peptides, (v) sequential extraction of the peptides with ACN, (vi) MALDI-snapshot of the extracts and (vii) investigation of the extract containing the most discriminating features using high resolution mass spectrometry. With this approach we have been able to differentially cluster renal oncocytoma and chromophobe renal cell carcinoma and identified 18 proteins specific to chromophobe and seven unique to renal oncocytoma. Chromophobes express proteins associated with ATP function (ATP5I & 5E; VATE1 & G2; ADT2), glycolysis (PGK1) and neuromedin whilst oncocytomas express ATP5H, ATPA, DEPD7 and TRIPB thyroid receptor interacting protein.

Genetic variants and copy number changes in soft tissue leiomyosarcoma detected by targeted amplicon sequencing

Aims

Leiomyosarcomas (LMSs) occur in various tissues and harbour potential for metastases. The genomic landscape of LMS is poorly understood. In an effort to improve understanding of the LMS genome, we analysed 11 LMSs of somatic soft tissue including matching tissue of normal phenotype.

Methods

DNA derived from microdissected tumour domains and matching normal tissue underwent amplicon sequencing of 409 tumour suppressors and oncogenes using the Ion Torrent Comprehensive Cancer Panel.

Results

Genomic changes were heterogeneous with few recurrent abnormalities detected. Coding variants were identified in genes involved in signal transduction, transcriptional regulation and DNA repair. There were variants in several genes related to angiogenesis and GPR124 variants (TEM5) were confirmed by immunohistochemical analysis of a LMS tissue microarray. Surprisingly, there were shared coding variants in tumour and corresponding normal tissue.

Conclusions

LMSs are a very heterogeneous population lacking recurrent somatic abnormalities. The presence of damaging mutations in normal tissue may reflect either a germline predisposition or field effect rather than tissue contamination. Hopeful therapeutic targets appear to be those related to AKT/MTOR pathway.

Development of a Robust Ultrasonic-Based Sample Treatment To Unravel the Proteome of OCT-Embedded Solid Tumor Biopsies

An effective three-step proteomics workflow is proposed to overcome the pitfalls caused by polymers present in optimum cutting temperature (OCT)-embedded tissue during its preparation for mass spectrometry analysis. First, the OCT-embedded tissue biopsies are cleaned using ethanol and water in a sequential series of ultrasonic washes in an ultrasound bath (35 kHz ultrasonic frequency, 100% ultrasonic amplitude, 2 min of ultrasonic duty time). Second, a fast ultrasonic-assisted extraction of proteins is done using an ultrasonic probe (30 kHz ultrasonic frequency, 50% ultrasonic amplitude, 2 min of ultrasonic duty time, 1 mm diameter tip). Third, a rapid ultrasonic digestion of complex proteomes is performed using a microplate horn assembly device (20 kHz ultrasonic frequency, 25% ultrasonic amplitude, 4 min of ultrasonic duty time). As a proof of concept, the new workflow was applied to human normal and tumor kidney biopsies including chromophobe renal cell carcinomas (chRCCs) and renal oncocytomas (ROs). A successful cluster of proteomics profiles was obtained comprising 511 and 172 unique proteins found in chRCC and RO samples, respectively. The new method provides high sample throughput and comprehensive protein recovery from OCT samples.

The HGF-MET signaling pathway is enriched in LUAC brain metastases

e20597

Background: Brain metastases occur in over 40% of non-small cell lung cancer (NSCLC) patients leading to a poor prognosis. c-Met (MET) is a receptor tyrosine kinase that upon binding hepatocyte growth factor (HGF), mediates proliferation, epithelial-mesenchymal transition (EMT), invasion, angiogenesis and metastasis. We have previously shown that the EMT transcription factor, TWIST1 is required for proliferation in MET driven NSCLC. Therefore, the HGF/MET/TWIST1 pathway may be a significant determinant of metastatic potential to the brain. Methods: We evaluated 125 lung adenocarcinoma (LUAC) brain metastases for MET amplification by FISH as well as other molecular alterations using targeted next generation sequencing in a subset of brain metastases (N = 74) and primary LUAC (N = 171) samples including 13 paired primary and brain sets. MET activation was examined in paired tumors using a HGF-MET proximity binding, dual-antibody assay (VeraTag; Monogram Biosciences). TWIST1 and EMT markers in the paired sets were measured by immunohistochemistry. Results: Compared to primary LUAC, we found that 17 pathogenic variants including TP53, SMAD4, RB1, RET, APC, ALK, FGFR3, EGFR, STK11 and MET alterations were significantly more common in LUAC brain metastases (adj. p values ≤ 0.02). Specifically, MET amplification was significantly enriched in LUAC brain metastases (23/125, 19%) compared to 2-4% in non-brain metastatic and primary sites. Among paired samples, 2/13 brain metastases had MET amplification that was not found in the primary tumor. MET mutations were also present in 16/74 brain cases (22%) compared to 9% (16/171) observed in the lung. VHL mutations were associated with MET altered cases compared to non- MET altered cases. MET expression was increased in the majority of brain metastases compared to the paired LUAC and there were 3 cases with brain specific MET activation. We found that TWIST1 was induced by HGF and determined response to MET TKIs in vitro. Among paired samples, TWIST1 was increased in brain metastases compared to primary LUAC in a subset of cases. Further analyses of TWIST1 and EMT markers is ongoing. Conclusions: Over a third of brain metastases have MET alterations compared to primary LUAC and may be responsive to MET inhibitors.

Expression profiles of immune-related genes are associated with neoadjuvant ipilimumab clinical benefit

Purpose: Patients with regionally advanced melanoma were treated with neoadjuvant ipilimumab in a previously reported study (PLOS One 2014). Gene expression profiles of tumors of treated patients were investigated for their association with immunotherapeutic benefit. Methods: Patients were treated with ipilimumab (10 mg/kg intravenously every 3 weeks × 2 doses) before and after surgery. Tumor specimens were obtained at baseline and at definitive surgery (weeks 6-8). Gene expression profiling was performed on the tumor biopsies of 27 patients. The primary endpoint was mRNA expression profiling using U133A 2.0 Affymetrix gene chips. Significance analysis of microarrays was performed to test the association of each gene with outcome. Pathway analysis was performed using Ingenuity Pathway Analysis software. The Benjamini and Hochberg method was used to adjust for multiple testing in the pathway analysis. Results: Pathway analysis identified biologically relevant pathways enriched with genes that are significantly associated with clinical outcome at baseline in relation to relapse-free survival (RFS) and disease non-progression (as assessed preoperatively at week 6) as well as early on-treatment (RFS and overall survival). The molecules and pathways that achieved differential expression of highest statistical significance were notably immune related. Association of the gene signature with clinical outcome overlapped between baseline and on-treatment specimens and across clinical endpoints tested. Conclusion: Gene expression profiling identified a signature reflecting an immune active and proinflammatory tumor microenvironment that derived clinical benefit from neoadjuvant ipilimumab at baseline and early on-treatment. These findings warrant further investigation in relation to ipilimumab and other immunotherapeutics.

Distinct human stem cell populations in small and large intestine

The intestine is composed of an epithelial layer containing rapidly proliferating cells that mature into two regions, the small and the large intestine. Although previous studies have identified stem cells as the cell-of-origin for intestinal epithelial cells, no studies have directly compared stem cells derived from these anatomically distinct regions. Here, we examine intrinsic differences between primary epithelial cells isolated from human fetal small and large intestine, after in vitro expansion, using the Wnt agonist R-spondin 2. We utilized flow cytometry, fluorescence-activated cell sorting, gene expression analysis and a three-dimensional in vitro differentiation assay to characterize their stem cell properties. We identified stem cell markers that separate subpopulations of colony-forming cells in the small and large intestine and revealed important differences in differentiation, proliferation and disease pathways using gene expression analysis. Single cells from small and large intestine cultures formed organoids that reflect the distinct cellular hierarchy found in vivo and respond differently to identical exogenous cues. Our characterization identified numerous differences between small and large intestine epithelial stem cells suggesting possible connections to intestinal disease.

Genome abnormalities precede prostate cancer and predict clinical relapse

The prediction of prostate cancer clinical outcome remains a major challenge after the diagnosis, even with improved early detection by prostate-specific antigen (PSA) monitoring. To evaluate whether copy number variation (CNV) of the genomes in prostate cancer tumor, in benign prostate tissues adjacent to the tumor (AT), and in the blood of patients with prostate cancer predicts biochemical (PSA) relapse and the kinetics of relapse, 241 samples (104 tumor, 49 matched AT, 85 matched blood, and 3 cell lines) were analyzed using Affymetrix SNP 6.0 chips. By using gene-specific CNV from tumor, the genome model correctly predicted 73% (receiver operating characteristic P = 0.003) cases for relapse and 75% (P < 0.001) cases for short PSA doubling time (PSADT, <4 months). The gene-specific CNV model from AT correctly predicted 67% (P = 0.041) cases for relapse and 77% (P = 0.015) cases for short PSADT. By using median-sized CNV from blood, the genome model correctly predicted 81% (P < 0.001) cases for relapse and 69% (P = 0.001) cases for short PSADT. By using median-sized CNV from tumor, the genome model correctly predicted 75% (P < 0.001) cases for relapse and 80% (P < 0.001) cases for short PSADT. For the first time, our analysis indicates that genomic abnormalities in either benign or malignant tissues are predictive of the clinical outcome of a malignancy.

In vitro transcription amplification and labeling methods contribute to the variability of gene expression profiling with DNA microarrays

The effect of different amplification and labeling methods on DNA microarray expression results has not been previously delineated. To analyze the variation associated with widely accepted T7-based RNA amplificationand labeling methods, aliquots of the Stratagene Human Universal Reference RNA were labeled using three eukaryotic target preparation methods followed by uniform replicate array hybridization (Affymetrix U95Av2). Method-dependent variability was observed in the yield and size distribution of labeled products, as well as in the gene expression results. A significant increase in short transcripts, when compared to unamplified mRNA, was observed in methods with long in vitro transcription reactions. Intramethod reproducibility showed correlation coefficients >0.99, whereas intermethod comparisons showed coefficients ranging from 0.94 to 0.98 and a nearly twofold increase in coefficient of variation. Fold amplification for each method positively correlated with the number of genes present. Our experiments uncovered two factors that introduced significant bias in gene expression data: the number of labeled nucleotides, which introduces sequence-dependent bias, and the length of the in vitro transcription reaction, which introduces transcript size-dependent bias. This study provides evidence that variability in expression data may be caused, in part, by differences in amplification and labeling protocols.

Control of attachment, morphology, and proliferation of skeletal myoblasts on silanized glass

Generating skeletal muscle in vitro is an attractive approach to overcome problems associated with autologous transfer of muscle and donor site morbidity during plastic surgery. Such tissue engineering requires application of biomaterials that selectively control the attachment, morphology, and proliferation of muscle progenitor ("satellite") cells. This study examined the initial attachment, morphological characteristics, and proliferative behavior of murine C2C12 myoblasts on glass substrata derivatized with self-assembled monolayers (SAMs) of the organosiloxanes N-(2-aminoethyl)(3-aminopropyl)trimethoxysilane (EDA) and tridecafluoro-1,1,2,2-tetrahydrooctyl-1-dimethylchlorosil ane (13F). The fraction of myoblasts resisting detachment upon rinsing was greater on EDA than on 13F. Application of a quantitative moments-based analysis of cell morphology demonstrated that projected area and two size-invariant metrics of shape (extension and dispersion) for these cells were greater for EDA than for 13F. Myoblasts also proliferated faster on EDA than on 13F. These data indicate that EDA-derivatized glass provides a superior substratum for myoblast culture compared to 13F-derivatized glass. Understanding myoblast behavior on these biomaterials that promotes contrasting cellular responses is the first step toward using patterned SAMs to control myotube alignment for tissue engineering skeletal muscle.