Brachytherapy-on-Chip: A Microfluidic Setup for In Vitro Interrogation of Hypoxic Spheroids

Abstract

Despite evidence of its advantages in many cancers, brachytherapy (BT) remains clinically underused and understudied in the pre-clinical setting due to a lack of versatile RT-compatible in vitro tools that can emulate the tumor microenvironment and radiobiology of various cancers. Microfluidic devices use conventional cell culture methods in 3D-tumor models, are radiocompatible and can integrate radiobioassays. However, they are seldom used in pre-clinical BT research. This project engineered the first microfluidic tool for in vitro testing of BT, with applications in translational radiobiology. PDMS microfluidic chips were engineered to grow and culture concentric rows of hypoxic spheroids, and to allow the insertion of a clinical iodine-125 BT seed at the center of the device. FaDu (hypopharyngeal squamous cell carcinoma), SK-LMS-1 (leiomyosarcoma) and HCT116 (colorectal carcinoma) cell lines were selected for their clinical relevance and ability to form spheroids. Presence of hypoxia in spheroids was assessed by immunofluorescence (IF) staining for hypoxic protein CAIX. On-chip dose distribution was calculated using clinical TG-43 parameters and compared to EBT-XD radiographic film. Target dose criteria was fixed at 8 Gy in the center of the first row of spheroids. Treatment response was quantified by DNA damages (γH2AX IF, comet assay) and cell survival (clonogenic assay). Response of hypoxic and normoxic regions of spheroids will be compared in IF. A total of 15 spheroids that are 750µm or larger can be cultured in our device, arranged as 5 rows of 3 spread from 1.5mm to 7.5mm away from a central iodine-125 seed. 48h after cell seeding, hypoxic cores were observed in spheroids derived from FaDu (50 ± 4% of cross-section, N = 3) and SK-LMS-1 (46 ± 4% of cross-section, N = 3) cells, results are pending for HCT116. TG-43 formula predicts that the center of each row receives 8 Gy, 2.6 Gy, 1.2 Gy, 0.7 Gy and 0.4 Gy respectively. Radiographic film analysis confirms on-chip TG-43 calculated doses (N = 5, R2 = 0.999). Similarly, tail moment from comet assays follows predicted dose trends (n>60, N = 3). There was no statistical difference between 8 Gy BT and 8 Gy GammaCell (8 Gy BT vs 8 Gy GammaCell, p = 0.8758), with a statistical difference between 8 Gy BT (8 Gy BT vs 2.6 Gy BT,1.2 Gy BT,0.7 Gy BT,0.4 Gy BT,0 Gy, p < 0.0001), 2.6 Gy BT (2.6 Gy BT vs 1.2 Gy BT,0.7 Gy BT,0.4 Gy BT,0 Gy, p < 0.05) or 8 Gy GammaCell (8 Gy GammaCell vs 2.6 Gy BT,1.2 Gy BT,0.7 Gy BT,0.4 Gy BT,0 Gy, p < 0.0001) and other lower BT doses. For the first time, brachytherapy can be easily integrated on-chip and its effects evaluated on relevant 3D-tumor models. Our system allows simultaneous quantification of BT efficacy on normoxic and hypoxic cells treated at various doses. On-chip combination of BT with antitumor drug will be explored in future work. We hope this device will serve as further proof of the potential of BT/RT-on-chip systems for better drug development, treatment planification and theranostics.