3D Tumor on a Chip with Embedded Vasculature Structure for Drug Screening Introduction

Cancer models, which are very important for developing diagnostics and treatments in cancer, include in vitro cell
culture and animal models,1-3 and have provided important advances over the past several decades. In vitro models
use mostly 2D platforms which are easy to utilize, but do have limitations due to a lack of mimicking the
physiological in vivo environment.4 There have been advances toward 3D systems in cancer research as scientists
strive to make the systems more physiologically relevant. For example, cancer cells can be cultured and
investigated in a 3D hydrogel scaffold thus mimicking in vivo structures. We developed a 3D collagen scaffold
embedded with tumor samples, specifically for drug screening and with potential for multiple tumor response
analysis and ease of imaging.
Materials and Methods
We fabricated microfluidic channels directly in collagen scaffolds through a combination of micromilling and
molding technologies. First, half-cylindrical microfluidic channels that were 500μm in diameter were micromilled
on a flat poly(methyl methacrylate) (PMMA) chip. Then these were replicated onto Polydimethylsiloxane
(PDMS) chips, and two PDMS chips with half channel pattern were combined face-to-face to create a cylindrical
channel pattern. Liquid gelatin was injected into the channel and solidified to make a sacrificial template, which
were then embedded in liquid collagen. After incubation at 37 oC, collagen was crosslinked and the gelatin
templates were melted and removed with a syringe, leaving channels in collagen scaffold.5
We then pre-seeded breast tumor fragment samples in liquid collagen, located in the crosslinked scaffold with
even interval. We fabricated two parallel channels for drug dosing (Figure 1A) with gelatin template and
generated a gradient of doxorubicin by injecting a determined dose into left channels.
Results and Discussion
Doxorubicin diffusion in collagen scaffold was tracked and characterized by detecting its autofluorescence
intensity. As shown in Figure 1B, doxorubicin diffused uniformly within 24 hours and formed a gradient of
concentration across the device. Next, tumor samples were placed in the device and treated with a dynamic
gradient of doxorubicin.
With precise simulation and control of the doxorubicin dose, we observed decreased tumor viability among the
tumor samples. We also compared the result with a traditional 96-well plates setup, in which tumors were directly
rinsed in media with a fixed drug dose. The results showed similar trend, yet a different lethal effect due to
different dosing pattern. Taken together, the 3D diffusion model was capable of testing tumor response to anticancer
drug in a dynamic mode.
Conclusions
We successfully developed a 3D collagen scaffold device with microfluidic channels for in vivo mimicking drug
screening applications. Our device enabled 3D embedded tumor samples to be treated with drug through ECM
diffusion from vessel mimicking channels, which simulated in vivo drug delivery process. We also applied a
simple 1D simulation to predict a drug diffusion profile and used it to determine local drug concentration for each
tumor sample. Besides in vivo mimic features, two other practical advantages of our device are: 1) Ease of
experimental manipulation. 2) Low drug consumption. As some drugs (especially new drugs in development)
could be rather expensive, our technique might be an economic choice.
References
1. B. S. Reddy, Lipids, 1992, 27, 807-813.
2. G. Y. Lee, P. A. Kenny, E. H. Lee and M. J. Bissell, Nature methods, 2007, 4, 359.
3. N. Kashaninejad, M. Nikmaneshi, H. Moghadas, A. Kiyoumarsi Oskouei, M. Rismanian, M. Barisam, M. Saidi and B.
Firoozabadi, Micromachines, 2016, 7, 130.
4. M. W. Tibbitt and K. S. Anseth, Biotechnology and bioengineering, 2009, 103, 655-663.
5. Wan, L., et al. "Mimicking Embedded Vasculature Structure for 3D Cancer on a Chip Approaches through
Micromilling." Scientific reports 7.1 (2017): 16724.