Biomedical Engineering Reference
In-Depth Information
increased bacterial adhesion. Increased bacterial
adhesion caused by increased roughness is a recurring
trend independent of bacterial strain and species
[61
e
69]
. Accumulation of bacteria around relatively
large surface features is an important observation of
bacterial adhesion to biomaterials. Investigations of
different topographies have shown that surface
features of the same order of magnitude as the
adhering bacteria promote adhesion
[69
e
72]
;thus,
individual
of orthopedic materials. Although the exact results
differed between strains, the propensity for bacteria
to adhere to injection-molded PEEK was similar to
microrough titanium in vitro. In addition, bacterial
adhesion to machined PEEK was generally higher
than to the titanium despite having similar roughness
values. This illustrates the significance of the specific
topography of machined PEEK for promoting
bacterial adhesion, as seen in
Fig. 8.7
.
To this end, any PEEK biomaterial topography
should be assessed postmanufacturing for features
that may promote bacterial adhesion and potentially
increase the risk of infection. Although the complete
elimination of topographical features may be bene-
ficial for reducing bacterial adhesion, a degree of
roughness is required to mediate eukaryotic cell
integration. Very smooth biomaterials typically
promote fibrous encapsulation to a greater extent
than the equivalent rough surface
[74
e
80]
.A
successful implant surface design should therefore
avoid surface features with dimensions similar to
bacteria, while maintaining a surface suitable for host
cell adhesion. Consequently, relatively smooth
biomaterials with nanotopographies may be an effi-
cient solution
[81]
. Nanoscale topographical features
may be suitable for increasing eukaryotic cell adhe-
sion through specific adhesion structures, yet too
small to promote bacterial adhesion. The literature
reports varying effects of nanotopographies, both
decreasing
[67]
and increasing bacterial adhesion
[82,83]
, and the exact role of nanoscale features may
depend on the method used to generate them.
Nanotopographies can be produced by a wide array
of techniques to give both defined and randomly
patterned surfaces
[81]
. Surface etching of injection-
molded PEEK by oxygen plasma treatment is one
such method that results in an increase in random
nanoscale features and micron-scale pits (
Fig. 8.8
)in
addition to an altered surface chemistry (discussed in
surface
structures with
dimensions
between 0.5 and 1
m should generally be avoided so
as to reduce bacterial adhesion.
The specific topography of a PEEK implant is
determined by the manufacturing method and any
postmanufacturing treatments applied. The most
common PEEK implant manufacturing processes are
machining, injection molding or a combination of the
two. These two manufacturing methods produce
distinct topographies (
Fig. 8.6
compared with stan-
dard microrough implant grade titanium). Machining
of PEEK implants results in a relatively rough
surface with nonuniform features, while injection
molding of PEEK produces a relatively smooth
topography (a reflection of the mold) with minimal
plateaus and ridges. Because of the distinct topog-
raphies produced by these manufacturing methods
[73]
, bacterial adhesion to machined PEEK and
injection-molded PEEK differs. Work conducted in
our lab with PEEK
e
OPTIMA
LT1 (PEEK) (Invibio
Biomaterical Solutions, UK) shows that a machined
topography leads to increased bacterial adhesion in
vitro, particularly around the larger surface features
present (
Fig. 8.7
A). The injection-molded topog-
raphy lacks these large features and therefore
bacteria adhere at a lower density in a more random
manner (
Fig. 8.7
B). Additionally, the adhesion of
bacteria to the PEEK surfaces was compared with
standard microrough implant grade titanium
surfaces, often considered to be the “gold standard”
m
Figure 8.5
Immunogold-labeled
S
.
aureus
illustrating the dis-
tribution of protein A, an
MSCRAMM common among
S
.
aureus
isolates. Reproduced
from Harris et al.
[58]
.
