Tissue engineering - Biomedical Engineering Desk Reference - page 628

150,000 patients die from the disease and its complica-
tions. Diabetes is characterized by pancreatic islet de-
struction leading to more or less complete loss of glucose
control. Tissue engineering approaches to treatment have
focused on transplanting functional pancreatic islets,
usually encapsulated to avoid immune reaction. Three
general approaches have been tested in animal experi-
ments. In the first, a tubular membrane was coiled in
a housing that contained islets. The membrane was
connected to a polymer graft that in turn connected the
device to blood vessels. This membrane had a 50-kDa
molecular mass cutoff, thereby allowing free diffusion of
glucose and insulin but blocking passage of antibodies and
lymphocytes. In pancreatectomized dogs treated with
this device, normoglycemia was maintained for more
than 150 days ( Sullivan et al. , 1991 ). In a second ap-
proach, hollow fibers containing rat islets were immobi-
lized in polysaccharide alginate. When the device was
placed intraperitoneally in diabetic mice, blood glucose
levels were lowered for more than 60 days and good
tissue biocompatibility was observed ( Lacy et al. , 1991 ).
Finally, islets have been enclosed in microcapsules com-
posed of alginate or polyacrylate. This method offers
a number of distinct advantages over the use of other
biohybrid devices, including greater surface-to-volume
ratio and ease of implantation (simple injection)
( Kin et al. , 2002 ; Lanza et al. , 1999 , 1995). All of these
transplantation strategies require a large, reliable source
of donor islets. Porcine islets are used in many studies
and genetically engineered cells that overproduce insulin
are also under investigation ( Efrat, 1999 ).

Fig. 7.1.2-2 Histologic photomicrograph demonstrating viable
hepatic cells after 2 days under flow conditions (hematoxylin and
eosin; original magnification 300). (Reprinted with permission
from Kim, S. S., et al., 1998. Ann. Surg. 228: 8-13.)

on appropriate polymers can form tissues resembling
those in the natural organ and have shown evidence of
bile ducts and bilirubin removal ( Uyama et al. , 1993 ).
More recently, model systems in which the vascular ar-
chitecture is mimicked in the device have been tested
using three-dimensional printing, hepatocytes, and en-
dothelial cells ( Fig. 7.1.2-2 ; Kim et al. , 1998 ).
Four bioartificial liver devices have entered sustained
clinical trials. The device rely all on hollow-fiber mem-
branes to isolate hepatocytes from direct contact with
patient fluids. They differ in source and treatment of
hepatocytes prior to patient use and in the choice of
perfusate: plasma or whole blood. Three devices are
perfused with the patient's plasma. The HepatAssist is
filled with freshly thawed cryopreserved primary porcine
hepatocytes along with collagen-coated dextran beads for
cell attachment ( Chen et al. , 1997 ; Rozga et al. , 1993 ;
Watanabe et al. , 1997 ). The ELAD system uses a HepG2
human hepatocyte cell line that has been grown to con-
fluence in the extracellular space (Ellis et al. , 1996;
Sussman et al. , 1994 ). The Gerlach BELS run either with
human hepatocytes (if available) or with porcine primary
hepatocytes embedded in a collagen matrix in the extra-
luminal space ( Gerlach, 1997 ; Gerlach et al. , 1997 ). In
contrast, the bioartificial liver support system (BLSS) is
perfused with whole blood. This has the advantage that
a greater rate of blood concentration reduction and lower
endpoint blood concentration at equivalent perfusion
times is achieved compared to systems using plasma per-
fusion. The detoxification is performed with primary
porcine hepatocytes ( Mazariegos et al. ,2001 ; Patzer et al. ,
2002 , 1999).

Tubular structures
The current concept of using tubular structures of other
organs for reconstruction of bladder, ureter and urethra,
trachea, esophagus, intestine, and kidney represents
a major therapeutic improvement. A diseased esophagus,
for example, can be treated clinically with autografts
from the colon, stomach, skin, or jejunal segments.
However, such procedures carry a substantial risk of graft
necrosis, inadequate blood supply, infection, lack of
peristaltic activity, and other complications. Copolymer
tubes consisting of lactic and glycolic acid have been su-
tured into dogs after removal of esophageal segments,
over time resulting in coverage of the polymer with
connective tissue and epithelium ( Grower et al. , 1989 ).
Alternatively, elastin-based acellular aortic patches have
been successfully used in experimental esophagus injury
in the pig. While mucosal and submucosal coverage took
place within 3 weeks, the majority of the elastin-based
biomaterial degraded. However, the muscular layer did
not regenerate ( Kajitani et al. , 2001 ). In a similar ap-
proach fetal intestinal cells have been placed onto co-
polymer

Pancreas
Each year more than 700,000 new cases of diabetes are
diagnosed in the United States and approximately tubes and implanted in rats. Histological