Identifying mitigating strategies for endothelial cell dysfunction and hypertension in response to VEGF receptor inhibitors.

Vascular endothelial growth factor receptor inhibitors (VEGFRis) improve cancer survival but are associated with treatment-limiting hypertension, often attributed to endothelial cell (EC) dysfunction. Using phosphoproteomic profiling of VEGFRi-treated ECs, drugs were screened for mitigators of VEGFRi-induced EC dysfunction and validated in primary aortic ECs, mice, and canine cancer patients. VEGFRi treatment significantly raised systolic blood pressure (SBP) and increased markers of endothelial and renal dysfunction in mice and canine cancer patients. α-Adrenergic-antagonists were identified as drugs that most oppose the VEGFRi proteomic signature. Doxazosin, one such α-antagonist, prevented EC dysfunction in murine, canine, and human aortic ECs. In mice with sorafenib-induced-hypertension, doxazosin mitigated EC dysfunction but not hypertension or glomerular endotheliosis, while lisinopril mitigated hypertension and glomerular endotheliosis without impacting EC function. Hence, reversing EC dysfunction was insufficient to mitigate VEGFRi-induced-hypertension in this mouse model. Canine cancer patients with VEGFRi-induced-hypertension were randomized to doxazosin or lisinopril and both agents significantly decreased SBP. The canine clinical trial supports safety and efficacy of doxazosin and lisinopril as antihypertensives for VEGFRi-induced-hypertension and the potential of trials in canines with spontaneous cancer to accelerate translation. The overall findings demonstrate the utility of phosphoproteomics to identify EC-protective agents to mitigate cardio-oncology side effects.

DNA Methylation-Derived Immune Cell Proportions and Cancer Risk, Including Lung Cancer, in Black Participants.

Prior cohort studies assessing cancer risk based on immune cell subtype profiles have predominantly focused on White populations. This limitation obscures vital insights into how cancer risk varies across race. Immune cell subtype proportions were estimated using deconvolution based on leukocyte DNA methylation markers from blood samples collected at baseline on participants without cancer in the Atherosclerosis Risk in Communities (ARIC) Study. Over a mean of 17.5 years of follow-up, 668 incident cancers were diagnosed in 2,467 Black participants. Cox proportional hazards regression was used to examine immune cell subtype proportions and overall cancer incidence and site-specific incidence (lung, breast, and prostate cancers). Higher T regulatory cell proportions were associated with statistically significantly higher lung cancer risk (hazard ratio = 1.22, 95% confidence interval = 1.06-1.41 per percent increase). Increased memory B cell proportions were associated with significantly higher risk of prostate cancer (1.17, 1.04-1.33) and all cancers (1.13, 1.05-1.22). Increased CD8+ naïve cell proportions were associated with significantly lower risk of all cancers in participants ≥55 years (0.91, 0.83-0.98). Other immune cell subtypes did not display statistically significant associations with cancer risk. These results in Black participants align closely with prior findings in largely White populations. Findings from this study could help identify those at high cancer risk and outline risk stratifying to target patients for cancer screening, prevention, and other interventions. Further studies should assess these relationships in other cancer types, better elucidate the interplay of B cells in cancer risk, and identify biomarkers for personalized risk stratification.

Atherosclerotic Cardiovascular Disease Risk and Longitudinal Risk Factor Management Among Patients With Breast Cancer.

BACKGROUND: Cardiovascular disease is the leading cause of noncancer mortality for breast cancer survivors. Data are limited regarding patient-level atherosclerotic cardiovascular disease (ASCVD) risk estimation and preventive medication use. This study aimed to characterize ASCVD risk and longitudinal preventive medication use for a cohort of patients with nonmetastatic breast cancer. PATIENTS AND METHODS: This retrospective cohort study included 326 patients at an academic medical center in Boston, Massachusetts diagnosed with nonmetastatic breast cancer or ductal carcinoma in situ from January 2009 through December 2015. Patient demographics, clinical characteristics, laboratory studies, medication exposure, and incident cardiovascular outcomes were collected. Estimated 10-year ASCVD risk was calculated for all patients from nonlaboratory clinical parameters. RESULTS: Median follow up time was 6.5 years (IQR 5.0, 8.1). At cancer diagnosis, 23 patients (7.1%) had established ASCVD. Among those without ASCVD, 10-year estimated ASCVD risk was ≥20% for 77 patients (25.4%) and 7.5% to <20% for 114 patients (37.6%). Two-hundred and sixteen patients (66.3%) had an indication for lipid-lowering therapy at cancer diagnosis, 123 of whom (57.0%) received a statin during the study. Among 100 patients with ASCVD or estimated 10-year ASCVD risk ≥20%, 92 (92.0%) received an antihypertensive medication during the study. Clinic blood pressure >140/90 mmHg was observed in 33.0% to 55.6% of these patients at each follow up assessment. CONCLUSION: A majority of patients in this breast cancer cohort had an elevated risk of ASCVD at the time of cancer diagnosis. Modifiable ASCVD risk factors were frequently untreated or uncontrolled in the years following cancer treatment.

Predicting In-Hospital Mortality After Acute Myeloid Leukemia Therapy: Through Supervised Machine Learning Algorithms.

PURPOSE: Despite careful patient selection, induction chemotherapy for acute myeloid leukemia (AML) is associated with a considerable risk for treatment-related mortality (5%-20%). We evaluated machine learning (ML) algorithms trained using factors available at the time of admission for AML therapy to predict death during the hospitalization. METHODS: We included AML discharges with age > 17 years who received inpatient chemotherapy from State Inpatient Database from Arizona, Florida, New York, Maryland, Washington, and New Jersey for years 2008-2014. The primary objective was to predict inpatient mortality in patients undergoing chemotherapy using covariates present before initiation of chemotherapy. ML algorithms logistic regression (LR), decision tree, and random forest were compared. RESULTS: 29,613 hospitalizations for patients with AML were included in the analysis each with 4,177 features. The median age was 58.9 (18-101) years, 13,689 (53.7%) were male, and 20,203 (69%) were White. The mean time from admission to chemotherapy was 3 days (95% CI, 2.9 to 3.1), and 2,682 (9.1%) died during the hospitalization. Both LR and random forest models achieved an area under the curve (AUC) score of 0.78, whereas decision tree achieved an AUC of 0.70. The baseline LR model with age yielded an AUC of 0.62. To clinically balance and minimize false positives, we selected a decision threshold of 0.7 and at this threshold, 51 of our test set of 5,923 could have potentially averted treatment-related mortality. CONCLUSION: Using readily accessible variables, inpatient mortality of patients on track for chemotherapy to treat AML can be predicted through ML algorithms. The model also predicted inpatient mortality when tested on different data representations and paves the way for future research.

Discordance in Oncotype DX Breast Recurrence Score® Results for Bilateral Breast Cancer.

BACKGROUND: The Oncotype DX Breast Recurrence Score® assay is a clinically useful tool to determine the benefit of chemotherapy in the treatment of early-stage, hormone-receptor-positive breast cancer. Bilateral breast cancer (BBC) is found in ~ 5% of patients with breast cancer, and data regarding discordance of Oncotype DX results between BBC defined by current TAILORx subgroups are limited. Our goals are to study the rate of Oncotype DX discordance between BBC and investigate whether such differences can affect chemotherapy treatment discussions. METHODS: Patients with BBC were identified in US samples submitted to Genomic Health for 21-gene testing between January 2019 and July 2020. The risk categories were defined as 0-25 and 26-100 as well as 0-17, 18-30, and 31-100 for all patients. Subgroup analysis was also performed for node-negative women age ≤ 50 years with Recurrence Score results of 0-15, 16-20, 21-25, and 26-100. RESULTS: 944 BBC patients with known nodal status (702 node negative, 242 node positive) were identified and included. Among node-negative patients aged > 50 years, the rate of discordance in Recurrence Score by group (0-25 and 26-100) was 4.2% (n = 598). For node-negative patients aged ≤ 50 years, the risk group was discordant in < 3% when considering the risk grouping of 0-25 and 26-100. However, upon subgroup analysis based on TAILORx analysis, the rate of discordance was 48.1% in these younger patients (n = 104). CONCLUSIONS: This study shows that a clinically relevant rate of discordance in Oncotype DX results in patients with BBC may impact medical decision-making regarding chemotherapy.

Cost-Effectiveness Analysis: Lymph Node Transfer vs Lymphovenous Bypass for Breast Cancer-Related Lymphedema.

BACKGROUND: Lymph node transfer (LNT) and lymphovenous bypass (LVB) have been described as 2 major surgical options for patients with breast cancer-related lymphedema (BCRL) who have failed conservative therapy. The objective of our study was to perform a cost-effectiveness analysis comparing LNT and LVB for the treatment of BCRL. STUDY DESIGN: Rates of infection, lymph leak, and failure of LNT and LVB were obtained from a previously published meta-analysis. Failure of surgery was defined as the inability to cease compression therapy postoperatively. Procedural costs were calculated from Medicare reimbursement rates. Cost of conservative management of postoperative surgical site infection, lymph leak, and continued decongestive physiotherapy after failed surgery were obtained from literature review. Average utility scores for each health state were calculated using a visual analog scale survey, then converted to quality-adjusted life years (QALYs). A decision tree was constructed, and incremental cost-effectiveness ratio was assessed at $50,000/QALY. Deterministic and probabilistic sensitivity analyses were performed to evaluate the robustness of our findings. RESULTS: LNT was less costly ($22,492 vs $31,927) and more effective (31.82 QALY vs 29.24 QALY) than LVB. One-way (deterministic) sensitivity analysis demonstrated that LNT became cost-ineffective when its failure rate was more than 43.8%. LVB became more cost-effective than LNT when its failure rate was less than 21.4%. Probabilistic sensitivity analysis using Monte-Carlo simulation indicated that even with uncertainty present in the variables analyzed, the majority of simulations (97%) favored LNT as the more cost-effective strategy. CONCLUSIONS: LNT is a dominant, cost-effective strategy compared to LVB for the treatment of BCRL.

Comparative Effectiveness of Adjuvant Chemotherapy in Early-Stage Breast Cancer: A Network Meta-analysis.

BACKGROUND: There are several regimens recommended by the National Comprehensive Cancer Network (NCCN) for HER2-negative operable breast cancer. To our knowledge, no trials have yet been performed comparing these regimens head to head. We performed a network meta-analysis comparing the efficacy of NCCN-recommended chemotherapy regimens. METHODS: We searched Medline, Embase, Web of Science, the Cochrane Central Register of Controlled Trials, and World Health Organization (WHO) International Clinical Trials Registry Platform from inception to February 2020. We included randomized clinical trials comparing adjuvant regimens in predominantly node-positive operable breast cancer patients. We compared (1) DDACT, (2) TCx4 cycles, (3) TAC, and (4) ACWKT. Common comparators were (5) AC, (6) ACT, and (7) ACD. Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines were followed. The Cochrane risk of bias tool assessed quality of the studies. Odds ratios (ORs) were calculated as measures of treatment effects with AC as reference. We used Bayesian hierarchical random-effects models with noninformative priors for mixed multiple treatment comparisons. Effectiveness was estimated by disease-free and overall survival using ORs. Sensitivity analyses were performed. Safety outcomes included febrile neutropenia. RESULTS: We identified 7 randomized controlled trials with 16,332 patients. TC (odds ratio [95% confidence interval], 0.71 [0.36-1.40]), TAC (0.77 [0.37-1.57]), ACWKT (0.68 [0.34-1.38]), and DDACT (0.72 [0.35-1.42]) were similar for overall survival. TC (0.64 [0.36-1.14]), TAC (0.67 [0.39-1.15]), ACWKT (0.63 [0.37-1.07]), and DDACT (0.59 [0.35-1.01]) had similar disease-free survival benefit. With regard to toxicity, TAC (2.67 [0.30-21.04]) had the highest odds of febrile neutropenia. CONCLUSION: The current generation of regimens are similar in efficacy. Given the lower toxicity of TCx4 comparatively, it is an acceptable alternative for lower-risk early-stage HER2-negative breast cancers.

Assessment of health-related quality of life among adults hospitalized with sickle cell disease vaso-occlusive crisis.

Sickle cell disease (SCD) is characterized by painful vaso-occlusive crises (VOCs). Self-reported pain intensity is often assessed with the Numeric Rating Scale (NRS), whereas newer patient-reported outcome measures (PROMs) assess multidimensional pain in SCD. We describe pain experiences among hospitalized adults with VOCs, using 2 PROMs: the Patient-Reported Outcomes Measurement Information System (PROMIS) Global Health and the Adult Sickle Cell Quality of Life Measurement System (ASCQ-Me). Adults with SCD hospitalized with VOCs at 2 academic centers in Boston, Massachusetts, from April 2016 to October 2017 were eligible. Participants completed the NRS and PROMs at admission and 7 days postdischarge. PROM scores were described and compared with population norms. Length of stay (LOS) and 30-day readmission rates were assessed. Forty-two (96%) of 44 eligible patients consented and completed admission assessments. Mean age was 30.2 years (standard deviation, 9.1), 60% were women, 76% were non-Hispanic black, and 64% had hemoglobin SS. Twenty-seven participants (64%) completed postdischarge assessments. Sixty percent had ≥4 VOCs in the last year. Nearly all PROMIS Global Health and ASCQ-Me scores were worse than population norms. NRS and PROMIS Global Physical Health scores improved after discharge, the latter driven principally by improvements in pain. Overall median LOS was 7 days, and 30-day readmission rate was 40.5%. Administration of PROMs among adults with SCD hospitalized for VOCs is feasible and demonstrates participants experienced recurrent, prolonged, and severe VOCs. PROMIS Global and ASCQ-Me scores indicated substantial suffering, and the striking 30-day readmission rate highlights the vulnerability of these patients.

Synthesis and Evaluation of Agelastatin Derivatives as Potent Modulators for Cancer Invasion and Metastasis.

The synthesis of new agelastatin alkaloid derivatives and their anticancer evaluation in the context of the breast cancer microenvironment is described. A variety of N1-alkyl and C5-ether agelastatin derivatives were accessed via application of our strategy for convergent imidazolone synthesis from a common thioester along with appropriately substituted urea and alcohol components. These agelastatin derivatives were evaluated in our three-dimensional coculture assay for the effects of mammary fibroblasts on associated breast cancer cells. We have discovered that agelastatin alkaloids are potent modulators for cancer invasion and metastasis at noncytotoxic doses. Herein, we discuss the increased potency of (-)-agelastatin E as compared to (-)-agelastatin A in this capacity, in addition to identification of new agelastatin derivatives with activity that is statistically equivalent to (-)-agelastatin E. The chemistry described in this report provides a platform for the rapid synthesis of agelastatin derivatives with excellent potency (50-100 nM) as modulators for cancer invasion and metastasis.

Cost Implications of Comorbidity for Autologous Stem Cell Transplantation in Elderly Patients with Multiple Myeloma Using SEER-Medicare.

Comorbidity is more common in older patients and can increase the cost of care by increasing toxicity. Using the SEER-Medicare database from 2000 to 2007, we examined the costs and life-year benefit of Auto-HSCT for MM patients over the age of 65 by evaluating the difference over time relative to comorbidity burden. One hundred ten patients had an Auto-HSCT in the early time period (2000-2003) and 160 in the late time period (2004-2007). Patients were divided by a Charlson Comorbidity Index (CCI) of 0 or greater than 1 (CCI1+). Median overall survival was 53.5 months for the late time period patients compared to 40.3 months for the early time period patients (p = 0.031). Median costs for CCI0 versus CCI1+ in the early period were, respectively, $70,900 versus $72,000 (100 d); $86,100 versus $98,300 (1 yr); and $139,200 versus $195,300 (3 yrs). Median costs for late period were, respectively, $58,400 versus $60,400 (100 d); $86,300 versus $77,700 (1 yr); and $124,400 versus $110,900 (3 yrs). Comorbidity had a significant impact on survival and cost among early time period patients but not among late time period patients. Therefore, older patients with some comorbidities can be considered for Auto-HSCT depending on clinical circumstances.

Where is the leak in the pipeline? Investigating gender differences in academic promotion at an academic medical centre.

BACKGROUND: Women are still under-represented in the senior ranks of academic medicine. As local surveys represent a critical initial step in addressing the challenges of gender disparities in academic promotion within institutions, we surveyed faculty at an academic medical centre to identify factors to improve the academic advancement of women. METHODS: We conducted an electronic survey of all full-time faculty members in a Department of Medicine assessing academic rank and factors important in consideration for promotion. RESULTS: 106 faculty members (46 %) responded to the survey; 40 % of the respondents were women. There was a statistically significant gender gap in faculty rank (p = 0.002), with only 2 of 17 full professor positions occupied by women. Among faculty who had not yet requested promotion, women were more likely to report that they did not think an academic promotion would benefit them (69 vs. 32 % in men, p = 0.01), and to report a lack of encouragement for requesting promotion (50 vs. 29 %, p = 0.08). CONCLUSIONS: Targeting the perceived value of academic promotion among women faculty, increasing junior faculty mentorship and modifying annual review processes could address gender disparities in academic medicine ranks.

Breast Cancer-Associated Fibroblasts: Where We Are and Where We Need to Go.

Cancers are heterogeneous tissues comprised of multiple components, including tumor cells and microenvironment cells. The tumor microenvironment has a critical role in tumor progression. The tumor microenvironment is comprised of various cell types, including fibroblasts, macrophages and immune cells, as well as extracellular matrix and various cytokines and growth factors. Fibroblasts are the predominant cell type in the tumor microenvironment. However, neither the derivation of tissue-specific cancer-associated fibroblasts nor markers of tissue-specific cancer-associated fibroblasts are well defined. Despite these uncertainties it is increasingly apparent that cancer-associated fibroblasts have a crucial role in tumor progression. In breast cancer, there is evolving evidence showing that breast cancer-associated fibroblasts are actively involved in breast cancer initiation, proliferation, invasion and metastasis. Breast cancer-associated fibroblasts also play a critical role in metabolic reprogramming of the tumor microenvironment and therapy resistance. This review summarizes the current understanding of breast cancer-associated fibroblasts.

The fibroblast Tiam1-osteopontin pathway modulates breast cancer invasion and metastasis.

BACKGROUND: The tumor microenvironment has complex effects in cancer pathophysiology that are not fully understood. Most cancer therapies are directed against malignant cells specifically, leaving pro-malignant signals from the microenvironment unaddressed. Defining specific mechanisms by which the tumor microenvironment contributes to breast cancer metastasis may lead to new therapeutic approaches against advanced breast cancer. METHODS: We use a novel method for manipulating three-dimensional mixed cell co-cultures, along with studies in mouse xenograft models of human breast cancer and a histologic study of human breast cancer samples, to investigate how breast cancer-associated fibroblasts affect the malignant behaviors of breast cancer cells. RESULTS: Altering fibroblast Tiam1 expression induces changes in invasion, migration, epithelial-mesenchymal transition, and cancer stem cell characteristics in associated breast cancer cells. These changes are both dependent on fibroblast secretion of osteopontin and also long-lasting even after cancer cell dissociation from the fibroblasts, indicating a novel Tiam1-osteopontin pathway in breast cancer-associated fibroblasts. Notably, inhibition of fibroblast osteopontin with low doses of a novel small molecule prevents lung metastasis in a mouse model of human breast cancer metastasis. Moreover, fibroblast expression patterns of Tiam1 and osteopontin in human breast cancers show converse changes correlating with invasion, supporting the hypothesis that this pathway in tumor-associated fibroblasts regulates breast cancer invasiveness in human disease and is thus clinically relevant. CONCLUSIONS: These findings suggest a new therapeutic paradigm for preventing breast cancer metastasis. Pro-malignant signals from the tumor microenvironment with long-lasting effects on associated cancer cells may perpetuate the metastatic potential of developing cancers. Inhibition of these microenvironment signals represents a new therapeutic strategy against cancer metastasis that enables targeting of stromal cells with less genetic plasticity than associated cancer cells and opens new avenues for investigation of novel therapeutic targets and agents.

Cost-Effectiveness of Autologous Hematopoietic Stem Cell Transplantation for Elderly Patients with Multiple Myeloma using the Surveillance, Epidemiology, and End Results-Medicare Database.

In the past decade, the number of autologous hematopoietic stem cell transplants (Auto HSCT) for older patients with multiple myeloma (MM) has increased dramatically, as has the cost of transplantation. The cost-effectiveness of this modality in patients over age 65 is unclear. Using the Surveillance, Epidemiology, and End Results-Medicare database to create a propensity-score matched sample of patients over age 65 between 2000 and 2007, we compared the survival and cost for those who received Auto HSCT to those who did not undergo transplantation but survived at least 6 months after diagnosis, and we calculated an incremental cost-effectiveness ratio (ICER). Two hundred seventy patients underwent transplantation. Median overall survival from diagnosis in those who underwent transplantation was significantly longer than in patients who did not (58 months versus 37 months, P < .001). For patients living longer than 2 years, the median monthly cost during the first year was significantly different, but the middle and last year of life costs were similar. The median cost of the first 100 days after transplantation was $60,000 (range, $37,000 to $85,000). The resultant ICER was $72,852 per life-year gained. Survival after transplantation was comparable to that in those who underwent transplantation patients under 65 years and significantly longer than older patients who did not undergo transplantation. With an ICER less than $100,000/life-year gained, Auto HSCT is cost-effective when compared with nontransplantation care in the era of novel agents and should be considered, where clinically indicated, for patients over the age of 65.

Ultra-sensitive protein detection via Single Molecule Arrays towards early stage cancer monitoring.

The early diagnosis of cancers and continued monitoring of tumor growth would be greatly facilitated by the development of a blood-based, non-invasive, screening technique for early cancer detection. Current technologies for cancer screening and detection typically rely on imaging techniques or blood tests that are not accurate or sensitive enough to definitively diagnose cancer at its earliest stages or predict biologic outcomes. By utilizing Single Molecule Arrays (SiMoA), an ultra-sensitive enzyme-linked immunosorbent assay (ELISA) technique, we were able to measure increasing levels of prostate specific antigen (PSA) within murine serum over time, which we attribute to tumor development. The measured concentrations of PSA were well below the detectable limits of both a leading clinical diagnostic PSA ELISA assay as well as a commercial ultra-sensitive PSA assay. Our work benchmarks the role of SiMoA as a vital tool in monitoring previously non-detectable protein biomarkers in serum for early cancer detection and offers significant potential as a non-invasive platform for the monitoring of early stage cancer.

Anatomical localization of progenitor cells in human breast tissue reveals enrichment of uncommitted cells within immature lobules.

INTRODUCTION: Lineage tracing studies in mice have revealed the localization and existence of lineage-restricted mammary epithelial progenitor cells that functionally contribute to expansive growth during puberty and differentiation during pregnancy. However, extensive anatomical differences between mouse and human mammary tissues preclude the direct translation of rodent findings to the human breast. Therefore, here we characterize the mammary progenitor cell hierarchy and identify the anatomic location of progenitor cells within human breast tissues. METHODS: Mammary epithelial cells (MECs) were isolated from disease-free reduction mammoplasty tissues and assayed for stem/progenitor activity in vitro and in vivo. MECs were sorted and evaluated for growth on collagen and expression of lineages markers. Breast lobules were microdissected and individually characterized based on lineage markers and steroid receptor expression to identify the anatomic location of progenitor cells. Spanning-tree progression analysis of density-normalized events (SPADE) was used to identify the cellular hierarchy of MECs within lobules from high-dimensional cytometry data. RESULTS: Integrating multiple assays for progenitor activity, we identified the presence of luminal alveolar and basal ductal progenitors. Further, we show that Type I lobules of the human breast were the least mature, demonstrating an unrestricted pattern of expression of luminal and basal lineage markers. Consistent with this, SPADE analysis revealed that immature lobules were enriched for basal progenitor cells, while mature lobules consisted of increased hierarchal complexity of cells within the luminal lineages. CONCLUSIONS: These results reveal underlying differences in the human breast epithelial hierarchy and suggest that with increasing glandular maturity, the epithelial hierarchy also becomes more complex.

Isolation of mammary epithelial cells from three-dimensional mixed-cell spheroid co-culture.

While enormous efforts have gone into identifying signaling pathways and molecules involved in normal and malignant cell behaviors(1-2), much of this work has been done using classical two-dimensional cell culture models, which allow for easy cell manipulation. It has become clear that intracellular signaling pathways are affected by extracellular forces, including dimensionality and cell surface tension(3-4). Multiple approaches have been taken to develop three-dimensional models that more accurately represent biologic tissue architecture(3). While these models incorporate multi-dimensionality and architectural stresses, study of the consequent effects on cells is less facile than in two-dimensional tissue culture due to the limitations of the models and the difficulty in extracting cells for subsequent analysis. The important role of the microenvironment around tumors in tumorigenesis and tumor behavior is becoming increasingly recognized(4). Tumor stroma is composed of multiple cell types and extracellular molecules. During tumor development there are bidirectional signals between tumor cells and stromal cells(5). Although some factors participating in tumor-stroma co-evolution have been identified, there is still a need to develop simple techniques to systematically identify and study the full array of these signals(6). Fibroblasts are the most abundant cell type in normal or tumor-associated stromal tissues, and contribute to deposition and maintenance of basement membrane and paracrine growth factors(7). Many groups have used three dimensional culture systems to study the role of fibroblasts on various cellular functions, including tumor response to therapies, recruitment of immune cells, signaling molecules, proliferation, apoptosis, angiogenesis, and invasion(8-15). We have optimized a simple method for assessing the effects of mammary fibroblasts on mammary epithelial cells using a commercially available extracellular matrix model to create three-dimensional cultures of mixed cell populations (co-cultures)(16-22). With continued co-culture the cells form spheroids with the fibroblasts clustering in the interior and the epithelial cells largely on the exterior of the spheroids and forming multi-cellular projections into the matrix. Manipulation of the fibroblasts that leads to altered epithelial cell invasiveness can be readily quantified by changes in numbers and length of epithelial projections(23). Furthermore, we have devised a method for isolating epithelial cells out of three-dimensional co-culture that facilitates analysis of the effects of fibroblast exposure on epithelial behavior. We have found that the effects of co-culture persist for weeks after epithelial cell isolation, permitting ample time to perform multiple assays. This method is adaptable to cells of varying malignant potential and requires no specialized equipment. This technique allows for rapid evaluation of in vitro cell models under multiple conditions, and the corresponding results can be compared to in vivo animal tissue models as well as human tissue samples.