Quantification of spatial tumor heterogeneity in immunohistochemistry staining images.

MOTIVATION: Quantitative immunofluorescence is often used for immunohistochemistry quantification of proteins that serve as cancer biomarkers. Advanced image analysis systems for pathology allow capturing expression levels in each individual cell or subcellular compartment. However, only the mean signal intensity within the cancer tissue region of interest is usually considered as biomarker completely ignoring the issue of tumor heterogeneity. RESULTS: We propose using immunohistochemistry image-derived information on the spatial distribution of cellular signal intensity (CSI) of protein expression within the cancer cell population to quantify both mean expression level and tumor heterogeneity of CSI levels. We view CSI levels as marks in a marked point process of cancer cells in the tissue and define spatial indices based on conditional mean and conditional variance of the marked point process. The proposed methodology provides objective metrics of cell-to-cell heterogeneity in protein expressions that allow discriminating between different patterns of heterogeneity. The prognostic utility of new spatial indices is investigated and compared to the standard mean signal intensity biomarkers using the protein expressions in tissue microarrays incorporating tumor tissues from 1000+ breast cancer patients. AVAILABILITY AND IMPLEMENTATION: THE R CODE FOR COMPUTING THE PROPOSED SPATIAL INDICES IS INCLUDED AS SUPPLEMENTARY MATERIAL: . SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Modeling qRT-PCR dynamics with application to cancer biomarker quantification.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) is widely used for molecular diagnostics and evaluating prognosis in cancer. The utility of mRNA expression biomarkers relies heavily on the accuracy and precision of quantification, which is still challenging for low abundance transcripts. The critical step for quantification is accurate estimation of efficiency needed for computing a relative qRT-PCR expression. We propose a new approach to estimating qRT-PCR efficiency based on modeling dynamics of polymerase chain reaction amplification. In contrast, only models for fluorescence intensity as a function of polymerase chain reaction cycle have been used so far for quantification. The dynamics of qRT-PCR efficiency is modeled using an ordinary differential equation model, and the fitted ordinary differential equation model is used to obtain effective polymerase chain reaction efficiency estimates needed for efficiency-adjusted quantification. The proposed new qRT-PCR efficiency estimates were used to quantify GUCY2C (Guanylate Cyclase 2C) mRNA expression in the blood of colorectal cancer patients. Time to recurrence and GUCY2C expression ratios were analyzed in a joint model for survival and longitudinal outcomes. The joint model with GUCY2C quantified using the proposed polymerase chain reaction efficiency estimates provided clinically meaningful results for association between time to recurrence and longitudinal trends in GUCY2C expression.

Validation of tumor protein marker quantification by two independent automated immunofluorescence image analysis platforms.

Protein marker levels in formalin-fixed, paraffin-embedded tissue sections traditionally have been assayed by chromogenic immunohistochemistry and evaluated visually by pathologists. Pathologist scoring of chromogen staining intensity is subjective and generates low-resolution ordinal or nominal data rather than continuous data. Emerging digital pathology platforms now allow quantification of chromogen or fluorescence signals by computer-assisted image analysis, providing continuous immunohistochemistry values. Fluorescence immunohistochemistry offers greater dynamic signal range than chromogen immunohistochemistry, and combined with image analysis holds the promise of enhanced sensitivity and analytic resolution, and consequently more robust quantification. However, commercial fluorescence scanners and image analysis software differ in features and capabilities, and claims of objective quantitative immunohistochemistry are difficult to validate as pathologist scoring is subjective and there is no accepted gold standard. Here we provide the first side-by-side validation of two technologically distinct commercial fluorescence immunohistochemistry analysis platforms. We document highly consistent results by (1) concordance analysis of fluorescence immunohistochemistry values and (2) agreement in outcome predictions both for objective, data-driven cutpoint dichotomization with Kaplan-Meier analyses or employment of continuous marker values to compute receiver-operating curves. The two platforms examined rely on distinct fluorescence immunohistochemistry imaging hardware, microscopy vs line scanning, and functionally distinct image analysis software. Fluorescence immunohistochemistry values for nuclear-localized and tyrosine-phosphorylated Stat5a/b computed by each platform on a cohort of 323 breast cancer cases revealed high concordance after linear calibration, a finding confirmed on an independent 382 case cohort, with concordance correlation coefficients >0.98. Data-driven optimal cutpoints for outcome prediction by either platform were reciprocally applicable to the data derived by the alternate platform, identifying patients with low Nuc-pYStat5 at ~3.5-fold increased risk of disease progression. Our analyses identified two highly concordant fluorescence immunohistochemistry platforms that may serve as benchmarks for testing of other platforms, and low interoperator variability supports the implementation of objective tumor marker quantification in pathology laboratories.

Prolactin-Stat5 signaling in breast cancer is potently disrupted by acidosis within the tumor microenvironment.

INTRODUCTION: Emerging evidence in estrogen receptor-positive breast cancer supports the notion that prolactin-Stat5 signaling promotes survival and maintenance of differentiated luminal cells, and loss of nuclear tyrosine phosphorylated Stat5 (Nuc-pYStat5) in clinical breast cancer is associated with increased risk of antiestrogen therapy failure. However, the molecular mechanisms underlying loss of Nuc-pYStat5 in breast cancer remain poorly defined. METHODS: We investigated whether moderate extracellular acidosis of pH 6.5 to 6.9 frequently observed in breast cancer inhibits prolactin-Stat5 signaling, using in vitro and in vivo experimental approaches combined with quantitative immunofluorescence protein analyses to interrogate archival breast cancer specimens. RESULTS: Moderate acidosis at pH 6.8 potently disrupted signaling by receptors for prolactin but not epidermal growth factor, oncostatin M, IGF1, FGF or growth hormone. In breast cancer specimens there was mutually exclusive expression of Nuc-pYStat5 and GLUT1, a glucose transporter upregulated in glycolysis-dependent carcinoma cells and an indirect marker of lactacidosis. Mutually exclusive expression of GLUT1 and Nuc-pYStat5 occurred globally or regionally within tumors, consistent with global or regional acidosis. All prolactin-induced signals and transcripts were suppressed by acidosis, and the acidosis effect was rapid and immediately reversible, supporting a mechanism of acidosis disruption of prolactin binding to receptor. T47D breast cancer xenotransplants in mice displayed variable acidosis (pH 6.5 to 6.9) and tumor regions with elevated GLUT1 displayed resistance to exogenous prolactin despite unaltered levels of prolactin receptors and Stat5. CONCLUSIONS: Moderate extracellular acidosis effectively blocks prolactin signaling in breast cancer. We propose that acidosis-induced prolactin resistance represents a previously unrecognized mechanism by which breast cancer cells may escape homeostatic control.

Low levels of Stat5a protein in breast cancer are associated with tumor progression and unfavorable clinical outcomes.

INTRODUCTION: Signal transducer and activator of transcripton-5a (Stat5a) and its close homologue, Stat5b, mediate key physiological effects of prolactin and growth hormone in mammary glands. In breast cancer, loss of nuclear localized and tyrosine phosphorylated Stat5a/b is associated with poor prognosis and increased risk of antiestrogen therapy failure. Here we quantify for the first time levels of Stat5a and Stat5b over breast cancer progression, and explore their potential association with clinical outcome. METHODS: Stat5a and Stat5b protein levels were quantified in situ in breast-cancer progression material. Stat5a and Stat5b transcript levels in breast cancer were correlated with clinical outcome in 936 patients. Stat5a protein was further quantified in four archival cohorts totaling 686 patients with clinical outcome data by using multivariate models. RESULTS: Protein levels of Stat5a but not Stat5b were reduced in primary breast cancer and lymph node metastases compared with normal epithelia. Low tumor levels of Stat5a but not Stat5b mRNA were associated with poor prognosis. Experimentally, only limited overlap between Stat5a- and Stat5b-modulated genes was found. In two cohorts of therapy-naïve, node-negative breast cancer patients, low nuclear Stat5a protein levels were an independent marker of poor prognosis. Multivariate analysis of two cohorts treated with antiestrogen monotherapy revealed that low nuclear Stat5a levels were associated with a more than fourfold risk of unfavorable outcome. CONCLUSIONS: Loss of Stat5a represents a new independent marker of poor prognosis in node-negative breast cancer and may be a predictor of response to antiestrogen therapy if validated in randomized clinical trials.

Epidermal growth factor receptor and insulinlike growth factor 1 receptor expression predict poor survival in pancreatic ductal adenocarcinoma.

BACKGROUND: The aim of this study was to evaluate the expression of epidermal growth factor receptor (EGFR) and insulinlike growth factor 1 receptor (IGF-1R) proteins and IGF-1R gene copy numbers in pancreatic ductal adenocarcinoma in relation to patients' characteristics and prognosis. METHODS: Immunohistochemical staining was performed on formalin-fixed paraffin-embedded tissue derived from tumor specimens recovered during surgery. Slides were evaluated for membranous EGFR and IGF-1R staining using both the HercepTest and the semiquantitative H-score systems. Chromogenic in situ hybridization was performed to quantify IGF-1R gene copy number. The primary outcome was the association between EGFR expression, IGF-1R expression-in both neoplastic epithelial and stromal cells-or IGF-1R gene copy number and overall survival. Secondary outcomes included associations between EFGR and IGF-1R expression and pathologic variables. RESULTS: A total of 105 patients were included. EGFR expression was present in 30.4% of cases and was associated with lymph node metastasis (P = .038). IGF-1R was overexpressed in 53% of tumors and correlated with higher tumor grade (P = .033). High membranous expression of EGFR (P < .001) and/or IGF-1R (P = .004), the cytoplasmic detection of EGFR (P = .027), and high expression levels of IGF-1R in the tumoral stroma (P < .001) were all associated with shorter overall survival, being significantly better in patients who simultaneously do not express membranous EGFR or stromal IGF-1R. CONCLUSIONS: EGFR and IGF-1R expression, in neoplastic and stromal cells, seems to be an important prognostic factor.

Loss of nuclear localized and tyrosine phosphorylated Stat5 in breast cancer predicts poor clinical outcome and increased risk of antiestrogen therapy failure.

PURPOSE: To investigate nuclear localized and tyrosine phosphorylated Stat5 (Nuc-pYStat5) as a marker of prognosis in node-negative breast cancer and as a predictor of response to antiestrogen therapy. PATIENTS AND METHODS: Levels of Nuc-pYStat5 were analyzed in five archival cohorts of breast cancer by traditional diaminobenzidine-chromogen immunostaining and pathologist scoring of whole tissue sections or by immunofluorescence and automated quantitative analysis (AQUA) of tissue microarrays. RESULTS: Nuc-pYStat5 was an independent prognostic marker as measured by cancer-specific survival (CSS) in patients with node-negative breast cancer who did not receive systemic adjuvant therapy, when adjusted for common pathology parameters in multivariate analyses both by standard chromogen detection with pathologist scoring of whole tissue sections (cohort I; n = 233) and quantitative immunofluorescence of a tissue microarray (cohort II; n = 291). Two distinct monoclonal antibodies gave concordant results. A progression array (cohort III; n = 180) revealed frequent loss of Nuc-pYStat5 in invasive carcinoma compared to normal breast epithelia or ductal carcinoma in situ, and general loss of Nuc-pYStat5 in lymph node metastases. In cohort IV (n = 221), loss of Nuc-pYStat5 was associated with increased risk of antiestrogen therapy failure as measured by univariate CSS and time to recurrence (TTR). More sensitive AQUA quantification of Nuc-pYStat5 in antiestrogen-treated patients (cohort V; n = 97) identified by multivariate analysis patients with low Nuc-pYStat5 at elevated risk for therapy failure (CSS hazard ratio [HR], 21.55; 95% CI, 5.61 to 82.77; P < .001; TTR HR, 7.30; 95% CI, 2.34 to 22.78; P = .001). CONCLUSION Nuc-pYStat5 is an independent prognostic marker in node-negative breast cancer. If confirmed in prospective studies, Nuc-pYStat5 may become a useful predictive marker of response to adjuvant hormone therapy.

Increased SIAH expression predicts ductal carcinoma in situ (DCIS) progression to invasive carcinoma.

Hyperactivated HER2/Neu/EGFR/RAS signaling is a major growth-promoting pathway known to drive cellular transformation and oncogenesis in breast cancers. HER2 amplification is detected in ~20% of all human breast cancer and is quite prevalent (up to 49%) in ductal carcinoma in situ (DCIS). The E3 ubiquitin ligase SIAH is considered a key downstream "gatekeeper" required for proper HER2/EGFR/RAS signal transduction. Formalin-fixed, paraffin-embedded resection specimens from 65 patients with DCIS treated with wide excision only were stained with an anti-SIAH antibody, and the percentage of tumor and normal adjacent tissue cells positive for SIAH nuclear staining were recorded. Statistical analysis was performed comparing SIAH staining in tumor cells to disease recurrence, histologic type, necrosis, hormone receptor status, and Her2/neu status, as well as nuclear grade. Correlation of SIAH expression in tumor cells with SIAH expression in normal adjacent tissue and age was also examined. Expression levels of SIAH in tumor cells was significantly higher in specimens from patients with recurrence (median = 19%) as compared to patients without recurrence (7%) (P < 0.001). There was also significantly increased SIAH expression in tumors with more aggressive features including comedo morphology (13.5% in comedo vs. 7% in other histologic types, P = 0.014). No significant association was observed between SIAH expression and estrogen receptor, progesterone receptor, and Her2/neu status. There was a significant correlation between SIAH expression in tumors and normal adjacent tissue (Spearman correlation = 0.58, P < 0.001) as well as between SIAH expression in normal adjacent tissue and patient age (Spearman correlation = -0.59, P < 0.001). No significant correlation was identified between patient age and SIAH expression in tumors (Spearman correlation = -0.23, P = 0.067). In conclusion, SIAH may represent a useful prognostic biomarker that predicts DCIS progression to invasive breast cancer.

HuR status is a powerful marker for prognosis and response to gemcitabine-based chemotherapy for resected pancreatic ductal adenocarcinoma patients.

BACKGROUND: Pancreatic ductal adenocarcinoma (PDA) is a devastating disease that killed nearly 38,000 people in the United States this past year. OBJECTIVE: Treatment of PDA typically includes surgery and/or chemotherapy with gemcitabine. No reliable biomarker exists for prognosis or response to chemotherapy. Two previously proposed prognostic markers, cyclooxygenase-2 (COX-2) and vascular endothelial growth factor (VEGF), are regulated by Hu protein antigen R (HuR), an mRNA binding protein that we have previously demonstrated to be a promising predictive marker of gemcitabine response. This study was designed to evaluate the clinical utility of HuR, COX-2, and VEGF as potential prognostic and predictive biomarkers for PDA. METHODS: A tissue microarray of 53 PDA specimens from patients who underwent potentially curative pancreatic resection was analyzed. HuR, COX-2, and VEGF status were correlated with clinicopathologic and survival data. We also performed ribonucleoprotein immunoprecipitation assays using an HuR antibody to assess VEGF and COX-2 mRNA binding to HuR in pancreatic cancer cells. RESULTS: Roughly 50% (27/53) of patients had high cytoplasmic HuR expression. These patients had worse pathologic features as assessed by T staging (P = 0.005). Only cytoplasmic HuR status correlated with tumor T staging, whereas VEGF (P = 1.0) and COX-2 (P = 0.39) expression did not correlate with T staging. Additionally, HuR status was an unprecedented positive predictive marker for overall survival in patients treated with gemcitabine, pushing median survival over 45 months in the high cytoplasmic HuR expressing patient population compared with less than 23 months in the low cytoplasmic HuR expressing patient group (P = 0.033 for log-rank test and P = 0.04 in a Cox regression model) for the low versus high cytoplasmic HuR expressing group. We also validated that mRNA transcripts for both VEGF and the gemcitabine metabolizing enzyme, deoxycytidine kinase, are specifically bound by HuR in pancreatic cancer cells. CONCLUSIONS: HuR is a useful prognostic biomarker for PDA patients as indicated by its association with higher tumor T stage. Additionally, HuR status is a robust predictor of outcome for patients with resected PDA in the setting of adjuvant gemcitabine therapy. Finally, HuR binds to VEGF mRNA implying that HuR, in part, regulates VEGF expression in PDA. This study supports the notion that HuR status should be used by clinicians for the individualized treatment of PDA in the future.

Stromal CD10 and SPARC expression in ductal carcinoma in situ (DCIS) patients predicts disease recurrence.

The current classification of ductal carcinoma in situ (DCIS) is based on nuclear grade, architectural differentiation and the presence of necrosis that does not adequately predict the likelihood of recurrence after breast conserving therapy; therefore, there is a critical need to identify novel predictors of DCIS progression. Ninety seven cases of DCIS were included in the study. CD10 and SPARC expression in tumor stroma was assessed by standard immunoperoxidase method with ani-CD 10 and anti-SPARC antibodies. The staining was scored semi-quantitatively as negative, weak or strong. Statistical analysis was performed using the Fisher's exact test. Multivariable analysis was conducted utilizing Exact Logistic Regression software (SAS 9.1 and LogExact). A significant association was observed between the recurrence status and time to recurrence with expression of CD10 (p < 0.001) and SPARC (p < 0.001). When combining both SPARC and CD10 expression there was a strong correlation with the shortest time to recurrence. Stromal CD10 and SPARC expression are new markers of an increased risk for DCIS recurrence, independent of commonly assessed clinical parameters. Thus, stromal CD10 and SPARC expression levels are promising markers of DCIS recurrence and warrant evaluation in larger prospective studies.