Network


Latest external collaboration on country level. Dive into details by clicking on the dots.

Hotspot


Dive into the research topics where Elin S. Agoston is active.

Publication


Featured researches published by Elin S. Agoston.


Histopathology | 2008

Tubal and ovarian pathways to pelvic epithelial cancer: a pathological perspective

Elke A. Jarboe; Ann K. Folkins; Ronny Drapkin; Tan A. Ince; Elin S. Agoston; Christopher P. Crum

Prolongation of ovarian epithelial cancer survival depends on early detection or improved responses to chemotherapy. Gains in either have been modest at best. Understanding the diverse pathogenesis of this disease is critical to early intervention or prevention. This review addresses six important variables, including (i) cell of origin, (ii) site of origin, (iii) initial genotoxic events, (iv) risks imposed by hereditary and other promoting conditions, (v) subsequent factors that promote different patterns of metastatic spread, and (vi) prospects for intervention. This review proposes two distinct pathways to pelvic epithelial cancer. The first initiates in ovarian surface epithelium (OSE), Mullerian inclusions or endometriosis in the ovary. The second arises from the endosalpinx and encompasses a subset of serous carcinomas. The serous carcinogenic sequence in the distal fallopian tube is described and contrasted with lower grade serous tumors based on tumour location, earliest genetic change and ability (or lack of) to undergo terminal (ciliated) differentiation. Ultimately, a clear understanding of tumour origin and the mechanism(s) leading to the earliest phases of the serous and endometrioid carcinogenic sequences may hold the greatest promise for designing prevention strategies and/or developing new therapies.


Cancer Research | 2011

Differentiation of NUT midline carcinoma by epigenomic reprogramming.

Brian E. Schwartz; Matthias D. Hofer; Madeleine E. Lemieux; Daniel E. Bauer; Michael J. Cameron; Nathan West; Elin S. Agoston; Nicolas Reynoird; Saadi Khochbin; Tan A. Ince; Amanda L. Christie; Katherine A. Janeway; Sara O. Vargas; Antonio R. Perez-Atayde; Stephen E. Sallan; Andrew L. Kung; James E. Bradner; Christopher A. French

NUT midline carcinoma (NMC) is a lethal pediatric tumor defined by the presence of BRD-NUT fusion proteins that arrest differentiation. Here we explore the mechanisms underlying the ability of BRD4-NUT to prevent squamous differentiation. In both gain-of and loss-of-expression assays, we find that expression of BRD4-NUT is associated with globally decreased histone acetylation and transcriptional repression. Bulk chromatin acetylation can be restored by treatment of NMC cells with histone deacetylase inhibitors (HDACi), engaging a program of squamous differentiation and arrested growth in vitro that closely mimics the effects of siRNA-mediated attenuation of BRD4-NUT expression. The potential therapeutic utility of HDACi differentiation therapy was established in three different NMC xenograft models, where it produced significant growth inhibition and a survival benefit. Based on these results and translational studies performed with patient-derived primary tumor cells, a child with NMC was treated with the FDA-approved HDAC inhibitor, vorinostat. An objective response was obtained after five weeks of therapy, as determined by positron emission tomography. These findings provide preclinical support for trials of HDACi in patients with NMC.


Nature Communications | 2015

Characterization of twenty-five ovarian tumour cell lines that phenocopy primary tumours

Tan A. Ince; Aurea D Sousa; Michelle Jones; J. Chuck Harrell; Elin S. Agoston; Marit Krohn; Laura M. Selfors; Wenbin Liu; Ken Chung Chen; Mao Yong; Peter Buchwald; Bin Wang; Katherine S. Hale; Evan Cohick; Petra A. Sergent; Abigail E. Witt; Zhanna Kozhekbaeva; Sizhen Gao; Agoston T. Agoston; Melissa A. Merritt; Rosemary Foster; Bo R. Rueda; Christopher P. Crum; Joan S. Brugge; Gordon B. Mills

Currently available human tumour cell line panels consist of a small number of lines in each lineage that generally fail to retain the phenotype of the original patient tumour. Here we develop a cell culture medium that enables us to routinely establish cell lines from diverse subtypes of human ovarian cancers with >95% efficiency. Importantly, the 25 new ovarian tumour cell lines described here retain the genomic landscape, histopathology and molecular features of the original tumours. Furthermore, the molecular profile and drug response of these cell lines correlate with distinct groups of primary tumours with different outcomes. Thus, tumour cell lines derived using this methodology represent a significantly improved platform to study human tumour pathophysiology and response to therapy.


American Journal of Clinical Pathology | 2010

Polymerase Chain Reaction Detection of HPV in Squamous Carcinoma of the Oropharynx

Elin S. Agoston; Stephen J. Robinson; Karishma Mehra; Chandler Birch; Dana R Semmel; Jelena Mirkovic; Robert I. Haddad; Marshall R. Posner; David Kindelberger; Jeffrey F. Krane; Joshua Brodsky; Christopher P. Crum

Human papillomavirus (HPV) testing is routinely performed on oropharyngeal carcinomas. We compared the Access Genetics (Minneapolis, MN) polymerase chain reaction (PCR) assay (AGPCR), DNA-DNA in situ hybridization (ISH; Ventana, Tucson, AZ), and HPV-16 E7 PCR amplification in consecutively accessioned oropharyngeal cancers. We tested 126 cases by both PCR methods; 102 were positive by either for a maximum positive rate (MPR) of 81.0%. Relative to the MPR, the sensitivities of AGPCR and E7 PCR were 90.2% and 72.5%, respectively. Of 17 AGPCR+ cases tested by ISH, 14/14 unequivocally positive/negative were concordant. All cases (97/97) positive by either PCR assay were positive for p16. There was no relationship between level of histologic differentiation and HPV status. ISH and AGPCR have comparable performance for the detection of HPV in oropharyngeal carcinomas. PCR is a suitable and economical assay that is comparable to ISH in sensitivity and may provide logistical advantages relative to ISH for assessing HPV status in oropharyngeal malignancies. However, it is imperative that appropriate sensitivity controls be in place for such assays.


PLOS ONE | 2013

Gene Expression Signature of Normal Cell-of-Origin Predicts Ovarian Tumor Outcomes

Melissa A. Merritt; Stefan Bentink; Matthew Schwede; Marcin P. Iwanicki; John Quackenbush; Terri Woo; Elin S. Agoston; Ferenc Reinhardt; Christopher P. Crum; Ross S. Berkowitz; Samuel C. Mok; Abigail E. Witt; Michelle Jones; Bin Wang; Tan A. Ince

The potential role of the cell-of-origin in determining the tumor phenotype has been raised, but not adequately examined. We hypothesized that distinct cells-of-origin may play a role in determining ovarian tumor phenotype and outcome. Here we describe a new cell culture medium for in vitro culture of paired normal human ovarian (OV) and fallopian tube (FT) epithelial cells from donors without cancer. While these cells have been cultured individually for short periods of time, to our knowledge this is the first long-term culture of both cell types from the same donors. Through analysis of the gene expression profiles of the cultured OV/FT cells we identified a normal cell-of-origin gene signature that classified primary ovarian cancers into OV-like and FT-like subgroups; this classification correlated with significant differences in clinical outcomes. The identification of a prognostically significant gene expression signature derived solely from normal untransformed cells is consistent with the hypothesis that the normal cell-of-origin may be a source of ovarian tumor heterogeneity and the associated differences in tumor outcome.


International Journal of Gynecological Pathology | 2009

P16 immunostaining patterns in microglandular hyperplasia of the cervix and their significance.

Michael H. Roh; Elin S. Agoston; Chandler Birch; Christopher P. Crum

P16 immunostaining is an important adjunct in the differential diagnosis of difficult squamous and glandular intraepithelial lesions of the cervix. However, unexpected staining of epithelium other than the target lesion can pose a problem in the interpretation. This study examined a common entity in the cervix, microglandular hyperplasia (MGH), that is associated with proliferations of both columnar and squamous epithelial cells—and ascertained the frequency of p16 staining, its pattern, and relationship to human papillomavirus. Fifty-seven cases of MGH were analyzed; 25 scored strongly immunopositive (44%). In 18, staining of the superficial columnar epithelium was patchy, involving 10% to 20% of cells on the surface; in 4 cases, 30% to 40% of cells; and in another 3, over 50% of the cells in a given area were strongly positive. Staining involved both nucleus and cytoplasm of columnar cells. P16 positivity did not colocalize with either cyclin E or MIB-1. Adjacent non–MGH-related columnar epithelium scored negative for p16. Of 25 p16-positive columnar epithelia analyzed, all were human papillomavirus -negative. In conclusion, benign columnar epithelium in the setting of MGH can be expected to stain strongly for p16. Practitioners should be aware of this when evaluating diagnostically difficult squamous or glandular epithelial changes occurring in the setting of MGH or when interpreting cytologic preparations stained with p16.


Cancer Research | 2018

Abstract 1054: Genomic representation of metastatic breast cancer patients by commercially available cell lines

Rick Nicoletti; Pamela Shaw; Naghmeh Salimi; Agoston T. Agoston; Elin S. Agoston

We probed the clinicopathologic context of commercially available cell lines using sequencing data from the recently released Metastatic Breast Cancer (MBC) Project. Additionally, we identified cohorts in the MBC with similar profiles to the Wood cell model we recently derived and investigated alignment between clinicopathologic characteristics of these patients and the patient from which the model was derived. Using a validated cancer panel, we analyzed 592 genes in the Wood cell model detecting mutations and CNAs. We then accessed mutation and CNA data from 1) the MBC Project for 103 primary and metastatic samples from 78 patients and 2) the Cancer Cell Line Encyclopedia (CCLE) database for 56 breast cancer cell lines. We performed unsupervised clustering on the 159 total specimens, incorporating both mutation and copy number data in the clustering analysis by labeling genes either altered or unaltered. MBC patient specimens were grouped based on the clustering results and assessed for trends across multiple clinicopathologic data categories. Specimens representing metastatic sites and the primary site from the same patient were both included in the analysis. Clustering analysis revealed three major groups representing high, medium and low alteration levels (means of 237, 104 and 29 panel genes altered). The High mutation load group had 34 cell lines and 6 patient samples, the Medium group had 15 cell lines and 71 patent samples, and the Low group had 7 cell lines (including Wood) and 26 patient samples. The 6 patient specimens in the High group represented 5.8% of the MBC data, yet were represented by a majority (60.7%) of CCLE breast cancer lines. All patients in the High group self-reported extraordinary response to treatment; the other two groups had even distribution. None of the specimens in the High group had evidence of DCIS reported in the path report, whereas observed DCIS was distributed equally outside of the High group. Expectedly, markers of aggressiveness of disease were lower in the Low group compared to the rest of the MBC patients, including histologic grade (p=0.00072) and Ki67 percentage (p=0.026). The Low group MBC samples were more likely to be HER2(-) (p=0.02) and ER(+) (p=0.037), consistent the Wood patient. The eight patients clustering most closely with the Wood cell model had low N-stage disease when compared to the rest (p=0.036) which was consistent with the Wood patient N0 status. CCLE breast cancer cell lines over-represent a small MBC patient subset with a very high load of genomic alterations that also trends toward exceptionally good response to treatment- consistent with the notion that cancer cell lines over-predict drug efficacy. Complete lack of DCIS in the High group is an interesting observation that may describe a disease progression resulting in absent or minimal DCIS precursor. As more MBC patient data are released, these analyses will be repeated to increase statistical power. Citation Format: Rick Nicoletti, Pamela Shaw, Naghmeh Salimi, Agoston T. Agoston, Elin S. Agoston. Genomic representation of metastatic breast cancer patients by commercially available cell lines [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1054.


Cancer Research | 2017

Abstract 821: Loss and retention of mutations in cell culture model systems

Pamela S. Shaw; Rick Nicoletti; Naghmeh Salimi; Helen L. Yang; Abigail E. Witt; Agoston T. Agoston; Elin S. Agoston

Although relevant, the genomic and proteomic alterations between original tumor tissue and newly established cell lines are frequently uncharacterized. In developing a novel cell model (Powder) of high grade serous carcinoma (HGSC), we followed the status of clinically relevant biomarkers to identify changes taking place between the tissue and cultured cell population. Understanding these changes enables proper utilization of new models and continued model optimization. Using a validated panel, we analyzed 592 genes in the Powder cell model and tumor from which it was derived, detecting several classes of genomic alterations. This included point mutations, indels, copy number variations, fusions, and variant transcripts. p53 mutation status of Powder was matched to a panel of previously reported HGSC cell lines. We further subdivided the panel by p53 protein expression level by reanalyzing previously published reverse phase protein array (RPPA) data. Significance of differential expression between groups was determined with a Welch9s t-test. A majority (14/21 or 67%) of the mutations detected in the HGSC tumor were conserved in the Powder cell model. Two additional mutations, undetected in the tumor, were found in the cell model. The tumor exhibited moderate to strong diffuse positive p53 immunostaining, and p53 mutations were measured in 79% of alleles. While the p53 mutation did not carry to the cell model, the cells were positive for CK7 and PAX8, consistent with ovarian cancer cells. In the panel of HGSC cell lines, we examined p53 mutation status and protein level, identifying 3 groups: (1) p53-mut/high protein, (2) p53-mut/low protein and (3) p53-WT/low protein. Powder identified with the p53-WT/low protein group. Previous work identifies the p53-WT/low protein group of cell models as part of a larger set of novel ovarian cell models with a stem-like molecular profile and higher drug resistance. Analysis of differential protein expression revealed upregulation in the p53 low protein group of both normal cell cycle genes and of notable cancer-related proteins. This included p21 and BAD, which are upregulated in cells with functional p53, eIF4E, a candidate cancer therapeutic target, and BOP1, which is disregulated in multiple solid cancers, including ovarian. These results underscore the need for further characterization of the Powder cell model to clarify its applications to specific research questions. Continued examination of Powder and other p53-WT cell models derived from HGSC tumors may uncover independent pathways driving disease. Additionally, understanding which mutations are underrepresented by cell models is necessary for ongoing methods development. Citation Format: Pamela S. Shaw, Rick Nicoletti, Naghmeh Salimi, Helen L. Yang, Abigail E. Witt, Agoston T. Agoston, Elin S. Agoston. Loss and retention of mutations in cell culture model systems [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 821. doi:10.1158/1538-7445.AM2017-821


Clinical Cancer Research | 2016

Abstract 33: Patient-specific cancer cell line derivation methodology results in genetic stability across 150 population doublings

Naghmeh Salimi; Jeffrey Kent; Alex Chao; Bryan E. Roberts; Agoston T. Agoston; Elin S. Agoston

Cancer cell lines have been, and will continue to be, important tools in oncology research, but there are major limitations in the fidelity of existing cell lines as a model of human cancers. Therefore, the aim of our study was to generate and characterize patient-specific cell lines from solid tumors that more faithfully recapitulate the primary tumor genetic and morphologic characteristics. We performed cell line derivation on a primary human breast adenocarcinoma (ER+/PR+/HER2-), a lung adenocarcinoma, a high-grade serous ovarian carcinoma, and an endometrioid ovarian carcinoma procured from a commercial tissue bank. Portions were snap-frozen for molecular analysis, and the remainder was minced and processed for cell culture. Cells were cultured and the resulting monolayer was stained by immunocytochemistry with a pan-keratin antibody for the presence of epithelial cells. Any contaminating fibroblasts were removed. We extracted DNA from the resulting breast and lung cancer cell lines at low, medium, and high passage numbers (approximately 20, 40, and 150 population doublings), as well as the corresponding primary tumor specimen, and a 250K SNP array was performed. These data were combined with that of nine breast cancer and 13 lung cancer cell lines from the GEO database, and subjected to genotyping analysis using the BLRMM algorithm in the Affymetrix Genotyping Console. The percent identity between the primary tumor tissue and the cell line was calculated from the number of identical SNP calls divided by the total number of SNP calls. Probesets resulting in “no call” genotyping results due to poor quality reads for either sample were not included in the analysis. For comparison, unrelated cancer cell line data of the same tumor type was included in the analysis. The cell lines reported here have grown for a minimum of 25 population doublings without evidence of cell culture crisis. The lung cancer, serous ovarian, and endometrioid ovarian cancer cell lines were positive for pan-keratin staining by immunocytochemistry. Both ovarian cancer subtypes exhibited tumorsphere formation under non-adherent culture conditions, whereas the lung cancer cell line formed loose aggregates. The percent identity between the primary carcinoma and cell line at 20, 40, and 150 population doublings was determined using a SNP call quality threshold of 0.1, resulting in an average 100K and 56K highest quality reads for breast and lung. The percent identity for the breast cancer cell line resulted in 98.7%, 97.4%, and 98.3% (mean +/- SD = 98.2% +/- 0.7%). The percent identity for the lung adenocarcinoma cell line at 20, 40, and 150 population doublings was 89.5%, 84.4%, and 87.2% (mean +/- SD = 87.0% +/- 2.5%). In comparison, the average % identity between unrelated existing breast cancer cell lines was 61.0% (+/- 3.1%) and between unrelated lung cancer cell lines was 58.7% (+/- 3.7%). These patient-specific breast and lung cancer cell lines exhibit genetically stable, extended growth from a primary tissue specimen for a minimum of 150 population doublings. We continue to measure extended growth in the ovarian cancer cell lines as well as a colon adenocarcinoma. Novel cell lines may be derived using this technology to represent specific patient groups and to serve the needs of researchers in precision medicine performing drug screening, target discovery, and the development of companion diagnostics. Citation Format: Naghmeh Salimi, Jeffrey D. Kent, Alex Chao, Bryan E. Roberts, Agoston Agoston, Elin S. Agoston. Patient-specific cancer cell line derivation methodology results in genetic stability across 150 population doublings. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Integrating Clinical Genomics and Cancer Therapy; Jun 13-16, 2015; Salt Lake City, UT. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(1_Suppl):Abstract nr 33.


Cancer Research | 2016

Abstract 4258: Multi-tumor cell culture medium supports a high take rate and improves culture growth rate in five tumor types

Elin S. Agoston; Marian Novak; Naghmeh Salimi; Alex Chao; Jeffrey Kent; Agoston T. Agoston; Ronny Drapkin

The efficiency of primary cell line derivation is inadequate, which presents a substantial hurdle in the study of many solid tumor types. We examined whether substitution of a novel multi-tumor cell culture medium into existing cell culture protocols may improve the efficiency of primary cell culture and cell line derivation. The multi-tumor medium RETM (Renaissance Essential Tumor Medium) was initially screened for the ability to achieve primary, extended, and continuous cultures from solid tumors of the breast, lung, colon, prostate and ovary (serous and endometrioid). A pan-cytokeratin antibody was then used to confirm epithelial origin of the expanded cell lines. Performance of RETM was compared to two established media types (RPMI, and Fallopian Tube Medium) in extended primary culture of five high-grade serous carcinoma (HGSC) tumor samples isolated from patients’ ascites. Immunofluorescent staining of Pax8, an important Mullerian lineage marker, was used to confirm the purity and the expected lineage identity of the resulting cell lines. To examine genetic drift, we performed SNP genotyping on the original tumor and the cultured cells in our extended primary cultures of HGSC and our continuous cultures from breast, lung, and colon. The resulting genotype was then compared to the original tumor and the percent matching genotype was determined. Of the six tumor types screened, five grew as long term cultures. Without additional optimization of methodology in HGSC, RETM supported a higher take rate (extended culture in 4 of 5 isolates versus 1 of 5 for established media types) as well as a 10-fold higher average growth rate in the same four cases (RETM doublings/day: 0.309 versus 0.030 for RPMI [p = 0.004] and 0.023 for FTM [p = 0.001]). In all cases, fibroblast overgrowth was ruled out by positive cytokeratin staining. The genetic analysis demonstrated a high degree of fidelity in genotypes between the patients’ tumors and the corresponding cell lines as follows: DF30 (HGSC at 23 doublings) 99.7%; DF68 (HGSC at 25 doublings) 95.5%; Wood (breast ductal and lobular carcinoma at 150 doublings) 98.3%; Jacket (lung adenocarcinoma at 150 doublings) 87.2%; and Ferry (colon adenocarcinoma at 30 doublings) 99.6%. We have shown for the first time that a single cell culture medium can support establishment of long-term cell lines established from a number of different tumor types including breast, lung, colon, endometrioid ovarian, and HGSC. Our results demonstrate that, compared to the current standard, RETM significantly improves both culture take rate and growth rate of the extended HGSC primary cultures. Together with the absence of cell culture crisis, the maintenance of genotype across a large number of population doublings indicate these cell lines are genetically stable and may be cultured to higher passages with reduced concern for genetic drift. Citation Format: Elin S. Agoston, Marian Novak, Naghmeh Salimi, Alex Chao, Jeffrey Kent, Agoston T. Agoston, Ronny Drapkin. Multi-tumor cell culture medium supports a high take rate and improves culture growth rate in five tumor types. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4258.

Collaboration


Dive into the Elin S. Agoston's collaboration.

Top Co-Authors

Avatar

Agoston T. Agoston

Brigham and Women's Hospital

View shared research outputs
Top Co-Authors

Avatar

Christopher P. Crum

Brigham and Women's Hospital

View shared research outputs
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar

Andrew L. Kung

Memorial Sloan Kettering Cancer Center

View shared research outputs
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Researchain Logo
Decentralizing Knowledge