📄 bioengineering

Cell-specific heteroclinic orbits govern interaction dynamics in inertial microfluidics for circulating tumour cell separation

This study demonstrates that incorporating cell-specific heteroclinic orbits and mechanical heterogeneity, rather than relying solely on size-based separation, is crucial for optimizing inertial microfluidic devices to minimize white blood cell carryover during circulating tumour cell isolation.

Original authors: Hay, R., Ghera, C., Zhou, J., Papautsky, I., Krueger, T., Owen, B.

Published 2026-09-17

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Hay, R., Ghera, C., Zhou, J., Papautsky, I., Krueger, T., Owen, B.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Every year, cancer claims millions of lives, often because it is discovered only after symptoms appear and the disease has advanced. To catch these illnesses earlier, doctors look for circulating tumour cells: tiny fragments of a tumour that break away and travel through the bloodstream. Finding these rare cells in a sample of blood could provide a crucial early warning, but the task is incredibly difficult. Blood is a crowded river of white blood cells, and the tumour cells are scarce, often appearing at concentrations of just five to one hundred cells in every millilitre. To separate them, scientists use microfluidic devices, which are tiny channels that guide cells using the physics of fluid flow. These devices rely on a simple principle: larger cells tend to move differently than smaller ones, allowing them to be sorted into different streams. However, this method is not perfect. Even with careful design, unwanted white blood cells often slip into the stream meant for the tumour cells, contaminating the sample and making analysis harder.

Researchers have long assumed that the key to better separation lies in the size difference between the cells. They believed that if they could simply make the tumour cells move faster toward the target stream than the white blood cells, the job would be done. But a new study suggests that this view is incomplete. By simulating the movement of cells in a straight microchannel, scientists discovered that the physical "squishiness" of the cells plays a massive role in how they interact. When a tumour cell and a white blood cell are close to each other, they do not just pass by independently; they influence one another. The study found that the way these cells deform as they flow changes the paths they take, and this interaction can drag unwanted white blood cells into the wrong stream, ruining the separation.

The researchers used a powerful computer simulation to watch how these cells behave in a channel that is only fifty micrometres high and one hundred and fifty micrometres wide. They modeled the cells as flexible, balloon-like spheres filled with fluid, similar to how real cells behave. In their simulations, they placed one tumour cell and eight white blood cells into the channel and watched them travel downstream. They tested different levels of stiffness, or deformability, for the tumour cells, ranging from very soft to quite stiff, while keeping the white blood cells relatively stiff. The goal was to see if the tumour cell's flexibility changed how the white blood cells moved.

The results showed that the presence of a tumour cell significantly altered the path of the white blood cells. When the tumour cell was very soft, the white blood cells stayed mostly on their own paths. But when the tumour cell was stiffer, the white blood cells were more likely to be pulled off course. The researchers observed two distinct types of interactions. In a "short" interaction, the cells would pass each other quickly, causing a brief jolt that might push a white blood cell slightly toward the target stream or back toward the wall. These quick encounters were frequent but did not cause a major shift in the overall flow.

However, a different scenario emerged when the tumour cell and a white blood cell had similar stiffness. In these cases, the two cells would get close and stay close for a long time, moving together at nearly the same speed. The researchers called this a "long" interaction. During these prolonged encounters, the white blood cell would drift away from its usual path and travel alongside the tumour cell, often ending up in the target stream where it did not belong. This happened because the two cells were following very similar routes through the channel, allowing them to stay in each other's wake for a long distance.

The study revealed that the distance between the natural paths of the two cell types is the deciding factor. When the tumour cell was very soft, its natural path was far away from the path of the stiffer white blood cells, so they rarely stayed close enough to interact for long. But when the tumour cell was stiffer, its path moved closer to the path of the white blood cells. This proximity made it much more likely for the two cells to get stuck in a long interaction, dragging the white blood cell into the target stream. The researchers found that the stiffer the tumour cell, the more likely it was to pull a white blood cell along with it.

This finding challenges the idea that size is the only thing that matters in these devices. The researchers suggest that the mechanical properties of the cells, specifically how easily they can be squashed, are just as important as their size. In the real world, cancer cells are not all the same; some are very soft, while others, especially those from early-stage cancers, can be quite stiff. If a device is designed only to separate based on size, it might fail to catch the stiff, early-stage tumour cells because they will interact too much with the white blood cells and get lost in the mix.

The team concluded that to build better devices, engineers need to think about more than just the size of the cells. They need to consider how the cells' stiffness affects their paths and how those paths bring the cells together. By understanding these interactions, designers might be able to adjust the flow or the shape of the channel to keep the paths of the tumour cells and white blood cells far enough apart that they do not get stuck together. This could lead to cleaner samples and a better chance of detecting cancer at its earliest, most treatable stages. The work does not offer a finished solution, but it provides a clear map of a hidden problem that has been overlooked, showing that the way cells move together is just as critical as the way they move alone.

Technical Summary: Cell-Specific Heteroclinic Orbits and Interaction Dynamics in Inertial Microfluidics for CTC Separation

Problem Statement
Circulating tumour cells (CTCs) are rare biomarkers for early cancer detection, yet their isolation from blood remains challenging due to their low concentration (5–100 cells/mL) and substantial heterogeneity in size and mechanical properties. Inertial microfluidics offers a label-free separation method based on cell size, where larger CTCs migrate laterally to a target stream faster than smaller white blood cells (WBCs). However, residual WBC carryover into the target stream limits separation purity. While conventional design relies on differences in single-cell migration rates, this approach fails to account for cell–cell interactions within the suspension. Specifically, it is unclear how the mechanical heterogeneity (deformability) of CTCs and WBCs influences their relative migration pathways and interaction dynamics, potentially leading to unexpected WBC migration toward the target stream.

Methodology
The authors employed 3D numerical simulations using a coupled lattice-Boltzmann-immersed-boundary-finite-element (LB-IBM-FEM) solver to model cell migration in a straight rectangular microchannel ( $30 \text{ mm} \times 150 \mu\text{m} \times 50 \mu\text{m}$ ).

Physical Model: Cells were modeled as deformable, neutrally buoyant spherical capsules with a hyperelastic membrane (Skalak model). The study simulated a dilute suspension (0.2% volume fraction) containing one CTC and eight WBCs.
Parameters: The Reynolds number was fixed at $Re=50$ . Cell deformability was quantified by the Laplace number ( $La$ ), varied across $[10, 20, 30, 50, 100]$ , where higher $La$ indicates lower deformability. CTC diameters were set to $17 \mu\text{m}$ and WBCs to $11 \mu\text{m}$ .
Analysis Framework: The study analyzed migration in terms of cell-specific heteroclinic orbits (HCOs)—the cross-sectional pathways cells follow toward equilibrium. The authors defined two interaction types based on proximity and duration:
- Short interactions: Brief overtaking events.
- Long interactions: Sustained proximity where cells maintain similar axial velocities.
Metric: A mean separation metric ( $\delta_{HCO}$ ) was developed to quantify the area between the HCOs of a CTC and a WBC, normalized by the CTC HCO perimeter.

Key Results

CTC Deformability Alters WBC Migration: The presence of a CTC significantly altered WBC migration compared to a WBC-only reference. Crucially, the deformability of the CTC dictated the extent of this alteration. When CTCs were less deformable ( $La_{CTC}=100$ ), WBCs entered the target stream significantly earlier (first reaching the centreline at $x/d_{CTC} \approx 500$ ) compared to when CTCs were more deformable ( $La_{CTC}=10$ , reaching the centreline at $x/d_{CTC} \approx 800$ ).
HCO Separation and Interaction Persistence: The study found that the relative separation between cell-specific HCOs ( $\delta_{HCO}$ $δ_{H C O}$ ) governs interaction frequency.
- Large $\delta_{HCO}$ : Occurs when CTCs are highly deformable and WBCs are rigid. This leads to a lower fraction of long interactions.
- Small $\delta_{HCO}$ : Occurs when CTCs are less deformable (closer in mechanical properties to WBCs). This reduces the separation between migration pathways, leading to a greater fraction of long interactions.
Impact of Interaction Types:
- Short interactions caused large, transient fluctuations in WBC migration rates but resulted in variable net migration.
- Long interactions caused sustained WBC migration toward the target stream. These interactions were the primary driver of increased WBC carryover.
Mechanism of Carryover: When CTC and WBC HCOs are closely aligned (small $\delta_{HCO}$ ), the cells experience similar axial flow velocities, allowing them to remain in close hydrodynamic proximity for extended distances. This proximity facilitates long interactions that displace WBCs from their isolated trajectories toward the target stream.

Significance and Claims
The paper argues that inertial microfluidic separation cannot be fully understood through single-cell migration rates alone. The authors claim that cell-specific heteroclinic orbits and the interactions they enable are critical design factors.

Mechanistic Insight: The study demonstrates that mechanical heterogeneity (specifically deformability) modifies the spatial relationship between CTC and WBC migration pathways.
Design Implication: To minimize WBC carryover, device design should aim to maximize the separation between cell-specific HCOs ( $\delta_{HCO}$ ). However, the authors caution that this must be balanced against maintaining reliable CTC entry into the target stream, as operating conditions that alter HCO separation may also affect HCO engagement and entry lengths.
Clinical Relevance: The findings are particularly significant for early-stage cancer detection, as early-stage CTCs often exhibit reduced deformability. This reduction could decrease HCO separation, increase long interactions, and consequently compromise separation efficiency and detection sensitivity.

Limitations
The authors note that the study is limited to a specific channel geometry and flow condition. The model assumes spherical, elastic capsules without internal structures, which simplifies the full mechanical heterogeneity of patient-derived cells. Additionally, the simulations utilized periodic boundary conditions and a higher CTC:WBC ratio than clinical samples to facilitate mechanistic analysis; thus, the interaction frequencies are relative comparisons rather than absolute predictions for physical devices. The specific interaction types identified were not experimentally validated in this work but are proposed as mechanistic predictions for future investigation.

More like this