Flexible regression models for ROC and risk analysis, with or without a gold standard.

A novel semiparametric regression model is developed for evaluating the covariate-specific accuracy of a continuous medical test or biomarker. Ideally, studies designed to estimate or compare medical test accuracy will use a separate, flawless gold-standard procedure to determine the true disease status of sampled individuals. We treat this as a special case of the more complicated and increasingly common scenario in which disease status is unknown because a gold-standard procedure does not exist or is too costly or invasive for widespread use. To compensate for missing data on disease status, covariate information is used to discriminate between diseased and healthy units. We thus model the probability of disease as a function of 'disease covariates'. In addition, we model test/biomarker outcome data to depend on 'test covariates', which provides researchers the opportunity to quantify the impact of covariates on the accuracy of a medical test. We further model the distributions of test outcomes using flexible semiparametric classes. An important new theoretical result demonstrating model identifiability under mild conditions is presented. The modeling framework can be used to obtain inferences about covariate-specific test accuracy and the probability of disease based on subject-specific disease and test covariate information. The value of the model is illustrated using multiple simulation studies and data on the age-adjusted ability of soluble epidermal growth factor receptor - a ubiquitous serum protein - to serve as a biomarker of lung cancer in men. SAS code for fitting the model is provided. Copyright © 2015 John Wiley & Sons, Ltd.

Ovarian cyst fluids are a cache of tumor biomarkers that include calgranulin A and calgranulin B isoforms.

SELDI-TOF MS analysis of cyst fluids identified 95 peaks that discriminate malignant, borderline, and benign ovarian tumors. Three prominent peaks, which correspond to calgranulin A (m/z 10847) and two isoforms of calgranulin B (m/z 12717 and 13294), have higher concentrations in borderline and malignant cyst fluids. Together, calgranulin A and B distinguish borderline and malignant tumors from benign tumors with 28.6% and 63.6% sensitivity for early stage disease, respectively, at 95% specificity and with 74.8% accuracy. Ovarian cyst fluids are useful for discovering discriminatory biomarkers, such as calgranulin, which may have utility for detecting, diagnosing, and biochemically classifying ovarian tumors.

Shedding of soluble epidermal growth factor receptor (sEGFR) is mediated by a metalloprotease/fibronectin/integrin axis and inhibited by cetuximab.

Soluble epidermal growth factor receptor (sEGFR) is a circulating serum biomarker in cancer patients. Recent studies suggest that baseline serum sEGFR concentrations may predict responsiveness to EGFR-targeted therapy. Here, we demonstrate that sEGFR is generated through proteolytic cleavage of a cell surface precursor of an alternately spliced EGF receptor isoform and that sEGFR binds to EGF with high affinity. Proteolytic cleavage is stimulated by an anti-α5/ß1 integrin antibody and 4-aminophenylmercuric acetate, and inhibited by fibronectin. Two FDA-approved therapeutic anti-EGFR antibodies also inhibit shedding of sEGFR, thus implicating the cell surface precursor of sEGFR as a competing target for anti-EGFR antibodies in human tissues. These observations parallel trastuzumab regulation of HER2 shedding and have implications for patient stratification in future clinical trials of EGFR-targeted antibodies.

Apolipoprotein concentrations are elevated in malignant ovarian cyst fluids suggesting that lipoprotein metabolism is dysregulated in epithelial ovarian cancer.

SELDI-TOF MS analysis of ovarian cyst fluids revealed that peaks m/z 8696 and 8825 discriminate malignant, borderline, and benign tumors. These peaks correspond to isoforms of apoA2. ELISA demonstrates that apoA1, A2, B, C2, C3, and E cyst fluid concentrations are uncorrelated and higher in malignant ovarian tumors, but only apoA2, apoE, and age are independent classifiers of malignant ovarian tumors, yielding 55.1% sensitivity, 95% specificity, and 88.1% accuracy to discern malignant from benign and borderline tumors. These data suggest that lipoprotein metabolism is dysregulated in ovarian cancer and that apoA2 and apoE warrant further investigation as ovarian tumor biomarkers.

A phase II evaluation of gefitinib in the treatment of persistent or recurrent endometrial cancer: a Gynecologic Oncology Group study.

BACKGROUND: A phase II trial was performed to evaluate the efficacy and safety of gefitinib in patients with persistent/recurrent endometrial cancer. METHODS: Women with histologically confirmed persistent/recurrent endometrial cancer were treated with 500mg oral gefitinib daily until progression or severe toxicity, with progression-free survival (PFS) at six months as the primary endpoint. Tumor expression of total epidermal growth factor receptor (EGFR), estrogen receptor (ER), progesterone receptor A (PRA) and B (PRB), Ki67, pEGFR and activated extracellular signal-regulated kinase (pERK) were examined pre- and post-treatment. EGFR was sequenced, and serum concentrations of soluble EGFR (sEGFR) at baseline also were examined. RESULTS: Of 29 patients enrolled, 26 were evaluable for efficacy and toxicity. Four patients experienced PFS ≥6 months, and one had a complete response which was not associated with an EGFR mutation. The concentration of sEGFR in pretreatment serum was positively correlated with overall survival (OS), but not with responsiveness to gefitinib in this small patient cohort. Expression of tumor biomarkers was not associated with PFS or OS. Co-expression of ER with PRA in primary and recurrent tumors, and pEGFR with pERK in primary tumors was observed. CONCLUSIONS: This treatment regimen was tolerable but lacked sufficient efficacy to warrant further evaluation in this setting. The possible association between serum sEGFR concentrations and OS, and temporal changes in expression of pEGFR and pERK and the documented CR of one patient are interesting and warrant additional investigation.

Soluble Human Epidermal Growth Factor Receptor 2 (sHER2) as a Potential Risk Assessment, Screening, and Diagnostic Biomarker of Lung Adenocarcinoma.

Lung cancer is the leading cause of cancer-related death in the United States. Here, we evaluated the potential clinical utility of soluble human epidermal growth factor receptor 2 (sHER2) for the risk assessment, screening, and diagnosis of non-small cell lung cancer (NSCLC) using an unmatched case-control study design. Serum sHER2 concentrations were measured by immunoassay in 244 primary NSCLC cases and 218 healthy controls. Wilcoxon rank-sum tests, logistic regression models, and receiver operating characteristic plots were used to assess whether sHER2 is associated with lung cancer. Median serum sHER2 concentrations are higher in patients with adenocarcinoma than squamous cell carcinoma regardless of gender, and sHER2 is a weak, independent biomarker of adenocarcinoma, but not of squamous cell carcinoma, adjusted for age and gender. The age-adjusted relative risk (odds) of adenocarcinoma is 3.95 (95% CI: 1.22, 12.81) and 7.93 (95% CI: 2.26, 27.82) greater for women and men with high sHER2 concentrations (≥6.60 ng/mL) vs. low sHER2 concentrations (≤1.85 ng/mL), respectively. When adjusted for each other, sHER2, age, and gender discern healthy controls from patients with primary adenocarcinomas of the lung with 85.9% accuracy. We conclude that even though serum sHER2 is not a strong, stand-alone discriminatory biomarker of adenocarcinoma, sHER2 may be a useful, independent covariate in multivariate risk assessment, screening, and diagnostic models of lung cancer.

Lin28 regulates BMP4 and functions with Oct4 to affect ovarian tumor microenvironment.

Emerging evidence suggests that the tumor microenvironment plays a critical role in regulating cancer stem cells (CSCs) and tumor progression through both autocrine and paracrine signaling. Elevated production of bone morphogenetic proteins (BMPs) from human ovarian cancer cells and stroma has been shown to increase CSC proliferation and tumor growth. Here, we report that Lin28, a stem cell factor, binds to BMP4 mRNA in epithelial ovarian carcinoma cells, thereby promoting BMP4 expression at the post-transcriptional level. As co-expression of Lin28 and Oct4 (another stem cell factor) has been implicated in ovarian cancer CSCs, we also determined that high levels of Lin28 are associated with an unfavorable prognosis when co-expressed with high levels of Oct4. Together, these findings uncover a new level of regulation of BMP4 expression and imply a novel Lin28/Oct4/BMP4-mediated mechanism of regulating ovarian tumor cell growth, thus holding potential for the development of new strategies for the diagnosis and treatment of ovarian cancer.

EGFR/HER-targeted therapeutics in ovarian cancer.

Despite decades of research and evolving treatment modalities, survival among patients with epithelial ovarian cancer has improved only incrementally. During this same period, the development of biologically targeted therapeutics has improved survival for patients with diverse malignancies. Many of these new drugs target the human epidermal growth factor receptor (EGFR/HER/ErbB) family of tyrosine kinases, which play a major role in the etiology and progression of many carcinomas, including epithelial ovarian cancer. While several HER-targeted therapeutics are US FDA approved for the treatment of various malignancies, none have gained approval for the treatment of ovarian cancer. Here, we review the published literature on HER-targeted therapeutics for the treatment of ovarian cancer, including novel HER-targeted therapeutics in various stages of clinical development, as well as the challenges that have limited the use of these inhibitors in clinical settings.

Identification of immunoreactive regions of homology between soluble epidermal growth factor receptor and α5-integrin.

Proteins encoded by the epidermal growth factor receptor (EGFR/HER1/ERBB1) gene are being studied as diagnostic, prognostic, and theragnostic biomarkers for numerous human cancers. The clinical application of these tissue/tumor biomarkers has been limited, in part, by discordant results observed for epidermal growth factor receptor (EGFR) expression using different immunological reagents. Previous studies have used EGFR-directed antibodies that cannot distinguish between full-length and soluble EGFR (sEGFR) expression. We have generated and characterized an anti-sEGFR polyclonal antiserum directed against a 31-mer peptide (residues 604-634) located within the unique 78-amino acid carboxy-terminal sequence of sEGFR. Here, we use this antibody to demonstrate that sEGFR is coexpressed with EGFR in a number of carcinoma-derived cell lines. In addition, we show that a second protein of ~140 kDa (p140) also is detected by this antibody. Rigorous biochemical characterization identifies this second protein to be α5-integrin. We show that a 26-amino acid peptide in the calf domain of α5-integrin (residues 710-735) is 35% identical in sequence with a 31-mer carboxy-terminal sEGFR peptide and exhibits an approximately 5-fold lower affinity for anti-sEGFR than the homologous 31-mer sEGFR peptide does. We conclude that the carboxy terminus of sEGFR and the calf-1 domain of α5-integrin share a region of sequence identity, which results in their mutual immunological reactivity with anti-sEGFR. We also demonstrate that anti-sEGFR promotes three-dimensional tissue cohesion and compaction in vitro, further suggesting a functional link between sEGFR and α5-integrin and a role of the calf-1 domain in cell adhesion. These results have implications for the study of both EGFR and sEGFR as cancer biomarkers and also provide new insight into the mechanisms of interaction between cell surface EGFR isoforms and integrins in complex processes such as cell adhesion and survival signaling.

Clinical implementation of soluble EGFR (sEGFR) as a theragnostic serum biomarker of breast, lung and ovarian cancer.

Signal transduction pathways regulated by the EGFR/ERBB/HER proto-oncogene family and receptor tyrosine kinases encoded by these genes are known to become dysregulated during cellular transformation and carcinogenesis. Consequently, biologically targeted antibodies and tyrosine kinase inhibitors directed toward EGFR/ErbB1/HER1 (eg, cetuximab, erlotinib and gefitinib) and ErbB2/HER2 (eg, trastuzumab), and more recently toward ErbB3/HER3 and ErbB4/HER4, are being investigated as therapeutic agents for treating patients with EGFR/ERBB/HER proto-oncogene-driven malignancies. The accurate selection of patients who will respond efficaciously to these agents a priori is a medical challenge. Understanding the clinical utility of soluble EGFR/ErbB/HER (ie, sEGFR/sErbB/sHER) isoforms, which are present in circulatory fluids, as theragnostic cancer biomarkers is an emerging area of contemporary biomedical investigation. This feature article reviews the literature regarding the clinical utility of serum sEGFR/sErbB1/sHER1 in breast, lung and ovarian cancer, and discusses the potential role of sEGFR in predicting and monitoring therapeutic responsiveness, as well as disease recurrence, and/or predicting disease outcome in patients treated with specific small-molecule, hormonal or biotherapeutic drug regimens. Well-designed translational research studies are needed to validate sEGFR as a theragnostic biomarker further and to achieve routine clinical implementation.

Trastuzumab-induced HER reprogramming in "resistant" breast carcinoma cells.

Although trastuzumab (Herceptin) is an important advance in the treatment of breast cancer, a significant proportion of patients do not respond to trastuzumab either alone or in combination with chemotherapy. In this study, we observe that epidermal growth factor receptor (EGFR) and HER3 expression is substantially increased after long-term trastuzumab exposure of HER2-positive breast carcinoma-derived cell lines that show primary resistance to trastuzumab. Furthermore, long-term trastuzumab exposure of trastuzumab-resistant cell lines induces de novo sensitivity to the EGFR-targeted agents gefitinib or cetuximab in two of three cell lines accompanied by increased EGFR expression. Together, these results indicate that primary trastuzumab resistance is not synonymous with lack of responsiveness to trastuzumab and, importantly, suggest that trastuzumab priming may sensitize trastuzumab-resistant tumors to other HER family-directed therapeutics.

Clinical implications of the ErbB/epidermal growth factor (EGF) receptor family and its ligands in ovarian cancer.

The ERBB or EGF receptor (EGFR) proto-oncogene family, which consists of four structurally-related transmembrane receptors (i.e., EGFR, ErbB2, ErbB3, and ErbB4), plays an etiological role in the molecular pathogenesis of cancer and is a key therapeutic target in many types of cancer, including ovarian cancer. These ErbB/EGF receptor tyrosine kinases play important physiologic roles in cell proliferation, survival, adhesion, motility, invasion, and angiogenesis. It is, therefore, not surprising that gene amplification, genetic mutation, and altered transcription/translation result in aberrant ErbB/EGF receptor expression and/or signal transduction, contributing to the development of malignant transformation. Clinically, the diagnostic, prognostic, and theragnostic significance of any single ErbB receptor and/or ErbB ligand is controversial, but generally, ErbB receptor overexpression has been correlated with poor prognosis and decreased therapeutic responsiveness in ovarian cancer patients. Thus, anticancer agents targeting ErbB/EGF receptors hold great promise for personalized cancer treatment. Yet, challenges remain in designing prospective clinical trials to assess the clinical utility of ErbB receptors and their ligands to diagnose cancer; to predict progression-free and overall survival, therapeutic responsiveness, and disease recurrence; and to monitor treatment responsiveness. Here, we review the tissue expression and serum biomarker studies that have evaluated the diagnostic, prognostic, and theragnostic utility of ErbB/EGF receptors, their circulating soluble isoforms (sEGFR/sErbBs), and their cognate ligands in ovarian cancer patients.

Decreased risk of bladder cancer in men treated with quinazoline-based α1-adrenoceptor antagonists.

Previous studies documented that human bladder cancer cells are sensitive to the apoptotic effects of quinazoline-derived α1-adrenoreceptor antagonists and bladder tumors exhibit reduced tissue vascularity in response to terazosin. More recent evidence suggests that exposure to quinazoline α1-adrenorecptor antagonists leads to a significant reduction in prostate cancer incidence. This retrospective observational cohort study was conducted to determine whether male patients treated with quinazoline α1-adrenoceptor antagonists for either benign prostate hyperplasia (BPH) or hypertension have a decreased risk of developing bladder cancer. Review of the medical records of all male patients enrolled at the Lexington Veterans Administration (VA) Medical Center identified men exposed to quinazoline-based α1-adrenoceptor antagonists (Jan 1, 1998-Dec 31, 2002) for either hypertension and/or benign prostate obstructive symptoms. The whole group of 27,138 male patients was linked to the Markey Cancer Center's Kentucky Cancer Registry (KCR), part of the NCI's Surveillance, Epidemiology, and End Results (SEER) Program, to identify all incident bladder cancer cases diagnosed in this population. Measures of disease incidence, relative risk, and attributable risk were calculated to compare the risk of developing bladder cancer for α1-blocker-exposed versus unexposed men. A two-by-two contingency table of α1-antagonist exposure versus bladder cancer diagnoses was constructed and the relative risk was calculated. Our analysis revealed a cumulative bladder cancer incidence of 0.24% among the α1-blocker-exposed men compared to 0.42% in the unexposed group. Thus, there was a risk difference of -0.0018, which indicates that 1.8 fewer bladder cancer cases developed per 1000 exposed men. Alternatively stated, 556 men would need to be treated with quinazoline α1-blockers to prevent one case of bladder cancer. Exposure to quinazoline α1-blockers thus may have prevented 7 to 8 bladder cancer cases among the 4173 treated men during the study period. The data yield an unadjusted risk ratio of 0.57 (95% CI: 0.30, 1.08) and therefore, men treated with α1-adrenoreceptor antagonists have a 43% lower relative risk of developing bladder cancer than unexposed men (p=0.083). Our inability to determine person-years at risk of developing bladder cancer for each unexposed control patient, was a limitation for calculating an incidence ratio and rate difference. These results offer an initial indication that exposure to doxazosin and terazosin decreases the incidence of bladder cancer. This is the first epidemiological evidence that the anti-tumor action of quinazoline-based α1-antagonists may potentially translate into a protective effect from bladder cancer development.

Effect of alpha1-adrenoceptor antagonist exposure on prostate cancer incidence: an observational cohort study.

PURPOSE: The quinazoline based alpha1-adrenoceptor antagonists doxazosin and terazosin suppress prostate tumor growth via the induction of apoptosis and decrease in tissue vascularity. To assess the effect of alpha1-blocker exposure on the incidence of prostate cancer we performed an exploratory, observational cohort study. MATERIALS AND METHODS: The medical records of all male patients enrolled at Lexington Veterans Affairs Medical Center were searched to identify men treated with quinazoline based alpha1-adrenoreceptor antagonists between January 1, 1998 and December 31, 2002 for hypertension and/or benign prostatic enlargement. Medical records were subsequently linked to the Markey Cancer Center Kentucky Cancer Registry, a statewide population based central cancer registry that is part of the National Cancer Institute Surveillance, Epidemiology and End Results Program, to identify all incident prostate cancer cases diagnosed. All newly diagnosed prostate cancer cases unexposed to alpha1-adrenoreceptor antagonists in the total male Veterans Affairs population during this period were also identified from the Kentucky Cancer Registry database. Measures of disease incidence, relative risk and attributable risk were calculated to compare the risk of prostate cancer in alpha1-blocker exposed vs unexposed men. Kaplan-Meier curves and Cox regression models were used to compare overall survival between alpha1-blocker exposed and unexposed prostate cancer cases. RESULTS: Our analysis revealed a cumulative incidence of 1.65% in alpha1-blocker exposed men compared to 2.41% in the unexposed group. These data yielded an unadjusted RR of 0.683 (95% CI 0.532, 0.876) and a risk difference of -0.0076, indicating that 7.6 fewer prostate cancer cases developed per 1,000 exposed men. Thus, exposure to quinazoline alpha1-blockers may have prevented 32 prostate cancer cases among the 4,070 treated men during the study period. Therefore, men exposed to quinazoline alpha1-adrenoceptor antagonists were at 1.46 times lower RR and 31.7% lower attributable risk for prostate cancer than unexposed men. There was no association between alpha1-adrenoceptor antagonist exposure and overall survival. CONCLUSIONS: These data suggest that exposure to quinazoline based alpha1-adrenoceptor antagonists significantly decreases the incidence of prostate cancer. This evidence suggests that the apoptotic and anti-angiogenic effects of these drugs may prevent the development of prostate cancer.

Soluble epidermal growth factor receptor acridinium-linked immunosorbent assay.

This chapter provides basic information for quantifying soluble epidermal growth factor receptor (sEGFR) isoforms in human sera using an acridinium-linked immunosorbent assay (ALISA) developed by Baron and coworkers. This ALISA has been shown to be specific for epidermal growth factor receptor (EGFR) and the isoforms of EGFR encoded by alternate transcripts (i.e., sEGFRs); it, therefore, does not detect other EGFR/ErbB receptor family members, i.e., ErbB2, ErbB3, or ErbB4. In addition, this ALISA recognizes EGFR and sEGFR isoforms that contain subdomains I-IV of the extracellular domain, but it does not recognize EGFR isoforms lacking subdomain IV. The ALISA described here also may be useful for determining EGFR and sEGFR concentrations in lysates of cultured cells, conditioned culture medium, or tissue/tumor specimens, as well as in other human body fluids such as serum, plasma, ascites, urine, saliva, and cystic fluids.

Serum soluble epidermal growth factor receptor concentrations decrease in postmenopausal metastatic breast cancer patients treated with letrozole.

Previous studies have implicated estrogen as a regulator of epidermal growth factor receptor (EGFR) expression in breast tumors. We therefore speculated that estrogen might modulate serologic soluble EGFR (sEGFR) concentrations in breast cancer patients. Accordingly, we measured serum sEGFR concentrations in postmenopausal women with metastatic breast cancer (MBC) treated with letrozole, an aromatase inhibitor that blocks estrogen synthesis. Serum specimens were obtained prior to and following 1 and 3 months of letrozole therapy. We report that sEGFR concentrations do not differ between MBC patients prior to letrozole treatment and age- and postmenopause-matched healthy women (P = 0.468). In contrast, however, sEGFR concentrations decreased significantly in 76% of MBC patients after both 1 month (P = 0.006) and 3 months (P = 0.003) of letrozole therapy versus pretreatment concentrations. Within the limitations of this study, we found no evidence for an association between pretreatment sEGFR concentrations or decreased treatment sEGFR concentrations and either progression-free or overall survival. Nonetheless, we conclude that future prospective studies are warranted to determine if baseline and/or longitudinal serum sEGFR concentrations may be useful for predicting disease progression and survival, and/or for monitoring responsiveness to aromatase inhibitors or other endocrine therapies in breast cancer patients.

Soluble epidermal growth factor receptor (sEGFR) [corrected] and cancer antigen 125 (CA125) as screening and diagnostic tests for epithelial ovarian cancer.

Epithelial ovarian cancer (EOC) is the leading cause of death among all gynecologic cancers in the United States. Because women who are diagnosed with early stage disease have a better prognosis than women diagnosed with late stage disease, early detection represents a potentially practical approach to reduce the mortality associated with EOC. Unfortunately, no single screening test has proven to be effective for this purpose, and a valid and feasible screening program to detect early stage EOC in the general population has not yet been devised. Consequently, research has focused on coupling two or more screening modalities to improve program validity and feasibility. Serum cancer antigen 125 (CA125) and a soluble isoform of the epidermal growth factor receptor (p110 sEGFR) have been studied individually as biomarkers of ovarian cancer. In this study, we compare serum CA125 levels and sEGFR concentrations in women with EOC to women with benign gynecologic conditions of ovarian and non-ovarian origin. We show that serum sEGFR concentrations are lower in patients with EOC than in women with benign gynecologic conditions, whereas serum CA125 levels are higher in patients to EOC compared with women with benign gynecologic conditions. These data also reveal that age and serum sEGFR concentrations modify the association between CA125 levels and EOC versus benign gynecologic disease. Hence, age- and sEGFR-dependent CA125 cutoff thresholds improve the ability of CA125 to discern EOC patients from women with benign ovarian tumors and non-ovarian gynecologic conditions. Our analyses show that parallel testing with fixed sEGFR and CA125 cutoff thresholds optimizes sensitivity to detect EOC, whereas serial testing with age- and sEGFR-dependent CA125 cutoff thresholds optimizes test specificity, and overall accuracy to discern patients with EOC from women with benign ovarian and non-ovarian gynecologic conditions. The combined use of serologic sEGFR and CA125, thus, has improved utility for screening and diagnosing EOC, which may increase the positive predictive value of a multimodal screening program that incorporates these biomarkers to detect and subsequently differentiate benign from malignant ovarian tumors.