anti-EGFR capture mitigates EMT- and chemoresistance-associated heterogeneity in a resistance-profiling CTC platform.

Capture and analysis of circulating tumor cells (CTCs) holds promise for diagnosing and guiding treatment of pancreatic cancer. To accurately monitor disease progression, capture platforms must be robust to processes that increase the phenotypic heterogeneity of CTCs. Most CTC-analysis technologies rely on the recognition of epithelial-specific markers for capture and identification, in particular the epithelial cell-adhesion molecule (EpCAM) and cytokeratin. As the epithelial-to-mesenchymal transition (EMT) and the acquisition of chemoresistance are both associated with loss of epithelial markers and characteristics, the effect of these processes on the expression of commonly used CTC markers, specifically EpCAM, EGFR and cytokeratin, requires further exploration. To determine this effect, we developed an in vitro model of EMT and acquired gemcitabine resistance in human pancreatic cancer cell lines. Using this model, we show that EMT-induction and acquired chemoresistance decrease EpCAM expression and microfluidic anti-EpCAM capture performance. Furthermore, we find that EGFR capture is more robust to these processes. By measuring the expression of known mediators of chemoresistance in captured cells using automated imaging and image processing, we demonstrate the ability to resistance-profile cells on-chip. We expect that this approach will allow for the development of improved non-invasive biomarkers of pancreatic cancer progression.

Comparison and optimization of machine learning methods for automated classification of circulating tumor cells.

Advances in rare cell capture technology have made possible the interrogation of circulating tumor cells (CTCs) captured from whole patient blood. However, locating captured cells in the device by manual counting bottlenecks data processing by being tedious (hours per sample) and compromises the results by being inconsistent and prone to user bias. Some recent work has been done to automate the cell location and classification process to address these problems, employing image processing and machine learning (ML) algorithms to locate and classify cells in fluorescent microscope images. However, the type of machine learning method used is a part of the design space that has not been thoroughly explored. Thus, we have trained four ML algorithms on three different datasets. The trained ML algorithms locate and classify thousands of possible cells in a few minutes rather than a few hours, representing an order of magnitude increase in processing speed. Furthermore, some algorithms have a significantly (P < 0.05) higher area under the receiver operating characteristic curve than do other algorithms. Additionally, significant (P < 0.05) losses to performance occur when training on cell lines and testing on CTCs (and vice versa), indicating the need to train on a system that is representative of future unlabeled data. Optimal algorithm selection depends on the peculiarities of the individual dataset, indicating the need of a careful comparison and optimization of algorithms for individual image classification tasks. © 2016 International Society for Advancement of Cytometry.

Detection of circulating pancreas epithelial cells in patients with pancreatic cystic lesions.

Hematogenous dissemination is thought to be a late event in cancer progression. We recently showed in a genetic model of pancreatic ductal adenocarcinoma that pancreas cells can be detected in the bloodstream before tumor formation. To confirm these findings in humans, we used microfluidic geometrically enhanced differential immunocapture to detect circulating pancreas epithelial cells in patient blood samples. We captured more than 3 circulating pancreas epithelial cells/mL in 7 of 21 (33%) patients with cystic lesions and no clinical diagnosis of cancer (Sendai criteria negative), 8 of 11 (73%) with pancreatic ductal adenocarcinoma, and in 0 of 19 patients without cysts or cancer (controls). These findings indicate that cancer cells are present in the circulation of patients before tumors are detected, which might be used in risk assessment.

Parametric control of collision rates and capture rates in geometrically enhanced differential immunocapture (GEDI) microfluidic devices for rare cell capture.

The enrichment and isolation of rare cells from complex samples, such as circulating tumor cells (CTCs) from whole blood, is an important engineering problem with widespread clinical applications. One approach uses a microfluidic obstacle array with an antibody surface functionalization to both guide cells into contact with the capture surface and to facilitate adhesion; geometrically enhanced differential immunocapture is a design strategy in which the array is designed to promote target cell­obstacle contact and minimize other interactions (Gleghorn et al. 2010; Kirby et al. 2012). We present a simulation that uses capture experiments in a simple Hele-Shaw geometry (Santana et al. 2012) to inform a target-cell-specific capture model that can predict capture probability in immunocapture microdevices of any arbitrary complex geometry. We show that capture performance is strongly dependent on the array geometry, and that it is possible to select an obstacle array geometry that maximizes capture efficiency (by creating combinations of frequent target cell­obstacle collisions and shear stress low enough to support capture), while simultaneously enhancing purity by minimizing nonspecific adhesion of both smaller contaminant cells (with infrequent cell­obstacle collisions) and larger contaminant cells (by focusing those collisions into regions of high shear stress).

The chemical neighborhood of cells in a diffusion-limited system.

Microorganisms follow us everywhere, and they will be essential to sustaining long-term human space exploration through applications such as vitamin synthesis, biomining, and more. Establishing a sustainable presence in space therefore requires that we better understand how stress due to the altered physical conditions of spaceflight affects our companion organisms. In microgravity environments such as orbital space stations, microorganisms likely experience the change in gravity primarily through changes in fluid mixing processes. Without sedimentation and density-driven convection, diffusion becomes the primary process governing the movement of growth substrates and wastes for microbial cells in suspension culture. Non-motile cells might therefore develop a substrate-deficient "zone of depletion" and experience stress due to starvation and/or waste build-up. This would in turn impact the concentration-dependent uptake rate of growth substrates and could be the cause of the altered growth rates previously observed in microorganisms in spaceflight and in ground-simulated microgravity. To better understand the extent of these concentration differences and their potential influence on substrate uptake rates, we used both an analytical solution and finite difference method to visualize concentration fields around individual cells. We modeled diffusion, using Fick's Second Law, and nutrient uptake, using Michaelis-Menten kinetics, and assessed how that distribution varies in systems with multiple cells and varied geometries. We determined the radius of the zone of depletion, within which cells had reduced the substrate concentration by 10%, to be 5.04 mm for an individual Escherichia coli cell in the conditions we simulated. However, we saw a synergistic effect with multiple cells near each other: multiple cells in close proximity decreased the surrounding concentration by almost 95% from the initial substrate concentration. Our calculations provide researchers an inside look at suspension culture behavior in the diffusion-limited environment of microgravity at the scale of individual cells.

Randomized, Noncomparative, Phase II Trial of Early Switch From Docetaxel to Cabazitaxel or Vice Versa, With Integrated Biomarker Analysis, in Men With Chemotherapy-Naïve, Metastatic, Castration-Resistant Prostate Cancer.

Purpose The TAXYNERGY trial ( ClinicalTrials.gov identifier: NCT01718353) evaluated clinical benefit from early taxane switch and circulating tumor cell (CTC) biomarkers to interrogate mechanisms of sensitivity or resistance to taxanes in men with chemotherapy-naïve, metastatic, castration-resistant prostate cancer. Patients and Methods Patients were randomly assigned 2:1 to docetaxel or cabazitaxel. Men who did not achieve ≥ 30% prostate-specific antigen (PSA) decline by cycle 4 (C4) switched taxane. The primary clinical endpoint was confirmed ≥ 50% PSA decline versus historical control (TAX327). The primary biomarker endpoint was analysis of post-treatment CTCs to confirm the hypothesis that clinical response was associated with taxane drug-target engagement, evidenced by decreased percent androgen receptor nuclear localization (%ARNL) and increased microtubule bundling. Results Sixty-three patients were randomly assigned to docetaxel (n = 41) or cabazitaxel (n = 22); 44.4% received prior potent androgen receptor-targeted therapy. Overall, 35 patients (55.6%) had confirmed ≥ 50% PSA responses, exceeding the historical control rate of 45.4% (TAX327). Of 61 treated patients, 33 (54.1%) had ≥ 30% PSA declines by C4 and did not switch taxane, 15 patients (24.6%) who did not achieve ≥ 30% PSA declines by C4 switched taxane, and 13 patients (21.3%) discontinued therapy before or at C4. Of patients switching taxane, 46.7% subsequently achieved ≥ 50% PSA decrease. In 26 CTC-evaluable patients, taxane-induced decrease in %ARNL (cycle 1 day 1 v cycle 1 day 8) was associated with a higher rate of ≥ 50% PSA decrease at C4 ( P = .009). Median composite progression-free survival was 9.1 months (95% CI, 4.9 to 11.7 months); median overall survival was not reached at 14 months. Common grade 3 or 4 adverse events included fatigue (13.1%) and febrile neutropenia (11.5%). Conclusion The early taxane switch strategy was associated with improved PSA response rates versus TAX327. Taxane-induced shifts in %ARNL may serve as an early biomarker of clinical benefit in patients treated with taxanes.

Automated electrorotation shows electrokinetic separation of pancreatic cancer cells is robust to acquired chemotherapy resistance, serum starvation, and EMT.

We used automated electrorotation to measure the cytoplasmic permittivity, cytoplasmic conductivity, and specific membrane capacitance of pancreatic cancer cells under environmental perturbation to evaluate the effects of serum starvation, epithelial-to-mesenchymal transition, and evolution of chemotherapy resistance which may be associated with the development and dissemination of cancer. First, we compared gemcitabine-resistant BxPC3 subclones with gemcitabine-naive parental cells. Second, we serum-starved BxPC3 and PANC-1 cells and compared them to untreated counterparts. Third, we induced the epithelial-to-mesenchymal transition in PANC-1 cells and compared them to untreated PANC-1 cells. We also measured the electrorotation spectra of white blood cells isolated from a healthy donor. The properties from fit electrorotation spectra were used to compute dielectrophoresis (DEP) spectra and crossover frequencies. For all three experiments, the median crossover frequency for both treated and untreated pancreatic cancer cells remained significantly lower than the median crossover frequency for white blood cells. The robustness of the crossover frequency to these treatments indicates that DEP is a promising technique for enhancing capture of circulating cancer cells.

Microfluidic immunocapture of circulating pancreatic cells using parallel EpCAM and MUC1 capture: characterization, optimization and downstream analysis.

We have developed and optimized a microfluidic device platform for the capture and analysis of circulating pancreatic cells (CPCs) and pancreatic circulating tumor cells (CTCs). Our platform uses parallel anti-EpCAM and cancer-specific mucin 1 (MUC1) immunocapture in a silicon microdevice. Using a combination of anti-EpCAM and anti-MUC1 capture in a single device, we are able to achieve efficient capture while extending immunocapture beyond single marker recognition. We also have detected a known oncogenic KRAS mutation in cells spiked in whole blood using immunocapture, RNA extraction, RT-PCR and Sanger sequencing. To allow for downstream single-cell genetic analysis, intact nuclei were released from captured cells by using targeted membrane lysis. We have developed a staining protocol for clinical samples, including standard CTC markers; DAPI, cytokeratin (CK) and CD45, and a novel marker of carcinogenesis in CPCs, mucin 4 (MUC4). We have also demonstrated a semi-automated approach to image analysis and CPC identification, suitable for clinical hypothesis generation. Initial results from immunocapture of a clinical pancreatic cancer patient sample show that parallel capture may capture more of the heterogeneity of the CPC population. With this platform, we aim to develop a diagnostic biomarker for early pancreatic carcinogenesis and patient risk stratification.