MMPs as diagnostic markers
A diagnostic marker is a molecule which can assist in diagnostic decision-making. Generally, the value of a diagnostic marker is determined by its ability to pick out patients with disease, the sensitivity, and its ability to leave out patients without disease, the specificity. In clinical practice,a number of different biomolecules are used in the diagnostics of cancer. Some of these can be detected in serum or urine and could therefore potentially be used as diagnostic markers in screening programs. However, to use a biomolecule as a tool to screen a general population, both the specificity and the sensitivity of the marker must be very high. If the marker molecule is used to select patients for further diagnostic procedures or to identify relapse in cancer patients, the sensitivity and specificity of the marker mole-
cule may be lower. Also, when the biomolecule is used as a tool to discriminate between various differential diagnoses, as in immunohistochemical analysis of biopsies, the requirement for sensitivity and specificity is not as high as if the molecules were used in a screening procedure.Increased levels of several MMPs have been reported in body fluids from cancer patients. This has raised the question if the MMP profile of blood, urine, or saliva can be used to diagnose cancer. This is a very appealing possibility as it opens for quick, non-invasive tests, which would be conve-
nient in screenings of large populations or patients with increased risk of cancer.
In a study of colorectal cancer (CRC), serum level of MMP-9 was measured in 300 symptomatic patients who were referred to a specialist colorectal clinic. The MMP-9 level was significantly higher in the serum from patients with malignant and premalignant lesions compared with the serum from patients with benign lesions. Dependent on the threshold value chosen, the test showed a sensitivity of up to 99 %
and a specificity of 63 % [36]. Most of the patients referred
for colonoscopic examinations with suspicion of cancer have
no malignant or premalignant lesions. If a quick, non-invasive
test could pick out those who would benefit from a
colonoscopic examination, time and resources would be spared.
The accuracy of serum MMP-9 levels as a test for CRC is
currently tested in a large prospective study of a primary care
population in the UK [37].
There are also studies showing that the presence of MMP-2
and MMP-9 in urine can be a valuable predictor of bladder
and prostate cancers [38, 39]. The diagnostic value was better
for bladder cancer than for prostate cancer, but in both cases,
the test was better at discriminating between patients with
malignant tumors and healthy individuals than between pa-
tients with malignant and benign tumors [38, 39]. Similarly,
there are quite a few studies assessing the diagnostic value of
MMP-2, MMP-7, MMP-9, and/or TIMP-1 or TIMP-2 in
tissue fluids like blood, urine, or saliva in various other can-
cers such as breast, gastric, esophageal, oral, brain, and renal
cancers [40–50]. All of these studies report that the level of
MMPs in the body fluid examined is increased in cancer
patients compared to healthy control individuals. Thus, the
MMP level in body fluids can often distinguish between
patients with malignant tumors and healthy individuals with
reasonably high sensitivity. In most cases, however, the tests
have a poor ability to separate between patients with malig-
nant tumors and patients with benign tumors or an inflamma-
tory disease. This may be because the increased levels of these
MMPs in cancer patients are more reflective of an inflamma-
tory reaction secondary to the tumors rather than the neoplas-
tic cells. Since the same MMPs are upregulated in many
different types of cancer as well as in many inflammatory
diseases, such tests cannot give a definite diagnosis. They may
however be valuable in screening patients at risk and to select
patients for further, more specific tests.
In clinical pathology, immunohistochemistry is often a
powerful tool to distinguish between benign and malignant
tumors and to distinguish between different types of cancers.
Antibodies against several dozens of cell-associated mole-
cules are routinely used. Despite the fact that many MMPs
are upregulated in various cancers, immunohistochemical
staining of MMPs has not become widely used in diagnostic
cancer pathology. One exception is MMP-11 which has been
shown to be a diagnostic marker discriminating between
dermatofibroma and dermatofibrosarcoma protuberance, a
benign and malignant skin tumor, respectively. It is often
difficult to distinguish between these two tumors based on
histopathological features alone. So far, antibodies against
CD34 and factor XIIIa have been used as a tool to separate
the two differential diagnoses, but recent studies have found
MMP-11 to be superior to both of them [51, 52].