Engineering a 3D in vitro model of cervical dysplasia to enable ethyl cellulose-ethanol ablation

Introduction:: When diagnosed in early stages cervical cancer has a 5-year survival rate of 92%¹. While current treatments for cervical dysplasia (precancerous stage) in Low Middle-Income Communities (LMICs) include cryotherapy and thermocoagulation, they are not easily accessible at the point of care due to costly consumables. To address this, we developed a low-cost accessible therapeutic based on ethanol ablation with the addition of a polymer ethyl cellulose (EC), which crosslinks to form a gel when injected into tissue; reduced leakage away from the injection site and improved efficacy in a small animal model of oral cancer². In order to evaluate EC-ethanol for cervical dysplasia treatment, most researchers use either xenograft flank mouse models with human cervical cancer cell lines or normal swine models which are not diseased³. However, neither widely used animal model recapitulates the microenvironment of cervical dysplasia, limiting the fundamental questions answerable with these models. Thus, we developed a 3D in vitro model of cervical dysplasia to evaluate delivery of EC-ethanol.

Materials and Methods:: Our 3D model will include a human cervical cancer cell line (SiHa) in the center, primary human fibroblasts surrounding it, and keratinocytes on top (Fig. 1A). The cells will be suspended in GelMa (8.7 % w/v) and photo-crosslinked with LAP (17 mg/mL). To ensure that the biophysical properties of our 3D model match the tissue we are trying to recapitulate, we characterized the mechanical properties of cervix tissue as well as our GelMa hydrogels. Using an AR-G2 rheometer (TA Instruments, New Castle, DE) equipped with a 25 mm standard steel parallel top plate (Fig. 1B), we conducted frequency sweeps from 0.1 to 10 rad/s within the linear viscoelastic region at a constant strain amplitude of 5% to obtain storage moduli (G') and loss moduli (G"). This was done for normal swine cervices (Animal Biotech Industries, Doylestown, PA), human cervices (University of Maryland School of Medicine, Baltimore, MD), and GelMa (8.7 % w/v, 17 mg/mL LAP).

Results, Conclusions, and Discussions:: Results and Discussion: Fig. 1C demonstrates that the viscoelastic properties of human and swine cervical tissue are in similar ranges. The 8.7% w/v GelMa had very similar storage moduli compared to the swine and human cervices, indicating that it captures the elastic contribution of cervical tissue’s resistance to deformation. The loss moduli of the GelMa was significantly lower, indicating that a wider range of GelMA hydrogel formulations should be explored to capture the elastic and the viscous contributions. Conclusions: We present here the first viscoelastic comparisons of swine and human cervices with a biosimilar hydrogel construct. Future work will include further tuning the hydrogel properties to match viscoelastic properties of cervical tissue. Additionally, we will administer EC-ethanol to the in vitro models to determine the optimal method to fully ablate lesions while minimizing fluid leakage into nearby normal tissue.

Acknowledgements (Optional): : This study was funded by the NIH (R00CA234455).


References (Optional): : 1. Sung H, A Cancer Journal for Clinicians. 2021
2. Morhard R, Sci Rep. 2017;7(1)
3. Larmour LI, Int J Gynecol Cancer. 2015;25(8):1345-52