Biomedical Engineering Reference
In-Depth Information
tissues. Indeed, inflammation has been shown to correspond to an increase of the endogenous fluo-
rescence of the tissue. It is presumably due to a strong metabolic activity of the cells involved in the
inflammatory response. An example is displayed in Figure 15.2a: we observe many circular structures
that exhibit a strong 2PEF signal and are less present in control lungs. These round cells have been
shown to correspond to macrophages using immunochemical labeling (Pena et al. 2007). Combined
SHG/2PEF microscopy therefore enables the discrimination of two major pathological processes, fibro-
sis, and inflammation that usually takes place simultaneously or sequentially in fibrotic pathologies. It
is exemplified in Figure 15.2c that shows an unstained section of murine lungs at 0, 3, 7, and 14 days
after bleomycin-injection. In that animal model, inflammation peaks at days 3-7, before fibrosis that is
detectable as early as day 3, and increases until day 14 (Pena et al. 2007).
15.3.2.5 Fibrosis/Sclerosis
As already noted above, SHG is highly specific to fibrillar collagens and gives no signal for the other
components of the ECM, including nonfibrillar collagens. As a consequence, SHG microscopy discrimi-
nates between
fibrosis
that is precisely defined as the accumulation of fibrillar collagen, and
sclerosis
that is the accumulation of nonfibrillar collagens and other matrix compounds. Both processes are
involved in fibrotic pathologies, but fibrosis onset is a crucial step in the pathology progression. Indeed,
fibrillar collagen is more resistant to proteolysis than nonfibrillar matrix and fibrosis regression is quite
unlikely. Fibrosis therefore appears to be a better predictor of severe pathological progression than scle-
rosis. For instance, interstitial fibrosis is considered as a strong indication for nephropathy progression
to end-stage renal disease (Nath 1992, Nicholson et al. 1996) although it shows poor functional out-
comes. Specific visualization and quantization of fibrosis is therefore of great interest for pathologists to
improve their diagnosis and to adjust the patient treatment. It is also crucial for fundamental studies to
better characterize the progression of fibrotic pathologies.
To summarize this section, SHG microscopy offers many advantages for fibrosis imaging, mainly
specificity and sensitivity to fibrillar collagen compared to other components of the matrix. It is there-
fore suitable for quantitative approaches as shown in the next section.
15.4. Fibrosis Quantitative Scoring Using SHG Microscopy
15.4.1 issues
The development of quantitative fibrosis indexes is a crucial issue for pathologists since conventional
techniques do not provide reliable and reproducible indexes. For instance, the interstitial fibrosis index
in the Banff classification that is used to assess renal allograft pathologies (Racusen et al. 1999, Solez
et al. 2007) shows poor reproducibility between different pathologists and embarrasses comparison of
clinical data from different centers (Marcussen et al. 1995, Furness et al. 2001, Gough et al. 2002, Seron
et al. 2002). Fibrosis scoring in liver affected by chronic hepatitis C also shows reproducibility limita-
tions (Bedossa et al. 1994).
The poor reproducibility and reliability of conventional techniques originate from two different limi-
tations. Conventional fibrosis scoring is based on semi-quantitative analysis of stained histological biop-
sies by trained pathologists. The first limitation is then related to the imaging technique itself, and the
second one is related to the image analysis. Regarding the imaging technique, the drawback of histologi-
cal staining is the lack of specificity to fibrillar collagen. Masson's trichrome, which is used for instance
in the Banff score of kidney allograft pathology, reveals in blue-green color all the components of the
ECM. Picrosirius red, which is recommended for instance for the METAVIR score of liver fibrosis in
hepatitis C, stains all collagen types and cannot discriminate between fibrillar and nonfibrillar collagens
(unless the biopsy is visualized by use of polarized microscopy which is not the usual technique in clini-
cal centers, see Section 5.1). Note that immunochemical labeling is a highly specific technique that could
be used to improve specificity of fibrosis scoring. However, this technique is not reproducible because it
