Biology Reference
In-Depth Information
almost invariably with fatal outcome. Although localized PC can be cured by radia-
tion therapy or surgery (radical prostatectomy), overtreatment of clinically
insignificant tumors is a major problem. Accordingly, there is a strong need for
improved diagnostic and prognostic biomarkers for this disease.
The development of miRNA biomarkers for PC is still at a relatively early stage
[ 140 ], although several studies have shown that miRNA expression profiles can be
used to distinguish PC from nonmalignant prostate as well as localized PC from
advanced metastatic PC (MPC) [ 26, 63, 107, 141- 147 ] . Inconsistencies have been
observed between studies, which at least in part can be explained by the use of different
profiling methodologies. Nevertheless, repeated findings in multiple studies show
upregulation of miR-375 and miR-200c and downregulation of miR-145, miR-205,
miR-221, and miR-222 in cancer vs . normal prostate. Molecular classifiers combin-
ing 2 (miR-205, miR-183), 3 (miR-143, miR-145, miR-375), or 54 different miRNAs
have been shown to correctly classify more than 75% of normal vs . PC samples
[ 63, 107, 147 ]. If these results are confirmed in future large-scale clinical studies, a
miRNA-based test could potentially improve the accuracy of PC diagnosis, e.g., as a
supplement to the currently used routine biomarker for PC detection, serum prostate-
specific antigen (PSA), which has highly suboptimal sensitivity and specificity.
In addition to their diagnostic value, miRNAs have shown potential for molecu-
lar classification of PC into clinically relevant subgroups for prognostic risk
stratification. For example, by analysis of miRNA expression profiles from 57 PC
tumors with or without perineural invasion (PNI), Pruitt et al . identi fi ed two sub-
groups of patient samples each associated with a distinct miRNA expression signa-
ture [ 148 ]. One subgroup included all non-PNI tumors as well as a subset of less
aggressive PNI tumors with lower stage and lower Gleason score, while the second
subgroup consisted of more aggressive PNI tumors characterized by higher stage
and higher Gleason score. Interestingly, clinically relevant clusters could not be
generated when the same analysis was performed on matching mRNA expression
data [ 148 ], again emphasizing miRNAs as promising biomarker candidates for PC.
Moreover, in a large miRNA profiling study of 102 clinical PC samples, Martens-
Uzunova et al . identified 80 miRNAs that were not only differentially expressed in
pairs of nonmalignant and tumor samples, but also separated clinically localized PC
(LPC) samples into two subgroups with distinct prognostic characteristics [ 147 ] .
The first group included 34 LPC samples clustered together with the nonmalignant
samples, while the second group included 16 LPC samples and clustered together
with advanced PC samples and lymph node metastases. Notably, the second group
was characterized by significantly higher risk of cancer-specific death. This finding
awaits confirmation in an independent patient cohort.
So far, few miRNAs have been shown to have prognostic value for PC in more
than one patient sample set. Schaefer et al . found that high miR-96 expression was
significantly associated with short recurrence-free survival after radical prostatec-
tomy in two independent PC patient sets [ 63 ]. In a separate study, Spahn et al . found
that low expression of miR-221 was an independent adverse prognostic factor for
recurrence-free survival after radical prostatectomy based on data from 92 PC
patients [ 145 ]. Consistent with this, Tong et al . identi fi ed a 16-miRNA signature,
including miR-211 , which correctly classified 75-85% of patients with and without
Search WWH ::




Custom Search