Biomedical Engineering Reference
In-Depth Information
The measurements of SHG
polarization,
on the other hand, can provide several pieces of infor-
mation regarding secondary and higher-order organization of collagen molecules and fibrils. First,
collagen SHG polarization may be used to indicate the angular orientation of the helix structure of
individual collagen molecules (i.e., the collagen pitch angle or helical angle) [36,37]. Polarization-
modulated SHG of collagen can also be employed to assess the angular distribution of collagen fibrils
[38-40]. These properties of collagen are significant for cancer research because an increasing num-
ber of studies support the idea that the changes in collagen's secondary and higher-order molecular
structure may predict the development or outcome of cancer. For example, the axial period of collagen
fibrils, which is in part a function of collagen's helical angle and higher molecular organization [41], is
demonstrably different in human breast cancer versus control tissue [42]. In addition, the proportions
of matrix molecules such as collagen I, III, fibrin, and fibronectin are frequently altered in reactive
tumor stroma [2,43,44]. The changes in the molecular collagen I:collagen III ratio in particular are
likely to alter several higher-order molecular characteristics of collagen fibrils [41,45], which may in
turn affect tumorigenesis or prognosis.
As such, in this chapter we highlight several optical methods and publications that have utilized the
coherent properties of SHG, specifically the scattering directionality and polarization, to investigate
collagen properties in tumor stroma.
17.1 Scattering Properties of SHG in tumors
17.1.1 Factors Affecting the
F
/
B
Signal
The SHG signal propagates both forward and backwards relative to the excitation laser axis. The
F
/
B
ratio from a distribution of scatterers, such as triple helices bundled into collagen fibrils, equals 1 for
small distributions of scatterers and increases as the size scale of the distribution surpasses the SHG
wavelength [1,25,46,47]. Therefore, the
F
/
B
ratio is closely related to the optically apparent thick-
ness of the fibril along the laser axis and within the focal volume [1,25-32], but is also influenced
by the excitation wavelength (as noted above), collagen packing density and order [47,48], and ionic
strength of the collagen's environment [27]. Briefly, the
F
/
B
ratio has been shown to be sensitive to
both collagen fibril diameter [31] and fibril orientation (angle) relative to the laser beam [32] (Figure
17.1), two factors that effectively determine the “optically apparent” fibril thickness in the
z
-axis.
Moreover, the
measured F
/
B
SHG ratio arises from the initial forward and backward SHG direction-
alities, as well as subsequent scattering of these signals (e.g., initially forward-directed SHG can scat-
ter to become backward-directed SHG), the latter of which is principally a function of tissue density
[46]. Similarly, biologic tissues will differentially attenuate the
F
and
B
SHG signals and thus, the
F
/
B
ratio is dependent on the tissue thickness, and in particular, the axial (
z
-axis) location of the focal
volume relative to the overall tissue depth [28,29,47-49]. Finally, reduction in ionic strength of the
collagen's environment dramatically increases the observed
F
/
B
(higher
F
and lower
B
), an effect that
is postulated to arise from swelling of the SHG-emitting collagen structures under hypo-osmotic
conditions [27].
As discussed elsewhere [28,46,48], we can also intuitively see that the contribution of some of these
factors (e.g., tissue depth and density) means that in the biologic tissue, the
final detected
(i.e., mea-
sured)
F
/
B
ratio generally differs from the scatterers'
initial emitted
or
original F
/
B
SHG directionality,
since subsequent scattering and absorption of the
initial
SHG signal with tissue depth and density will
alter the final measured SHG
F
/
B
. With calibration approaches, the
original
initial emitted
F
/
B
can be
inferred from the measured
F
/
B
[1,27], but it can only effectively be measured directly when the tissue
thickness is <1 mean free path (MFP) [28] (see discussions below for further details). Thus, the
detected
F
/
B
is dependent on both collagen structure
and
other tissue characteristics, whereas the
initial emit-
ted F
/
B
is primarily dependent on collagen structural characteristics. Herein, we will use
F
d
/
B
d
when
referring to the
final detected F
/
B
ratio measured in tissues, and
F
i
/
B
i
when referencing studies that have
