Development of a novel circulating model of CLL. (A) HUVEC cells were seeded into the hollow fibers of the bioreactor and exposed to physiologically relevant shear forces. Medium containing CLL cells was pumped around the system, permitting transient interaction with the endothelial layer. The hollow fibers in the bioreactor have pores through which CLL cells can migrate into the space outside fibers; the EVS is shown in the cross-sectional image of the cartridge. CLL cells were recovered from the circulating compartment and the EVS of the system via the access ports. (B) Photograph of the adapted bioreactor. (C) Scanning electron micrograph of HUVEC cells on the interior of the hollow fibers after 5 hours of alignment at 5 dynes/cm². The beginning of flattening and spreading of the cells is visible. (D) Scanning electron micrograph of HUVEC cells inside a hollow fiber after 6 hours of alignment under 10 dynes/cm², followed by 24 hours at 5 dynes/cm². The HUVEC cells showed increased spreading and flattening, resulting in coverage of the interior of the hollow fiber. (E) Scanning electron micrograph of a transverse section of a lined hollow fiber showing the wall of the fiber, with CLL cells visible in the inside of the fiber. Scale bars represent 10 microns in all scanning electron micrograph images. (F) Expression of endothelial cell markers was measured by flow cytometry on HUVEC cells grown in both noncirculating static tissue culture flasks and on cells recovered from the hollow fibers after alignment under shear force. Expression of VCAM-1 (P = .01), PECAM-1 (P = .002), and VEGFR2 (P = .05) were reduced in HUVEC cells under shear force. (G) CLL cell viability in circulating and static coculture with HUVEC cells. There was no significant difference between CLL viability in circulating culture and in static coculture with HUVEC cells. All culture conditions were supplemented with interleukin 4 (5 ng/mL).