Blood-based miRNAs as noninvasive diagnostic and surrogative biomarkers in colorectal cancer

Citation metadata

From: Expert Review of Molecular Diagnostics(Vol. 13, Issue 2)
Publisher: Expert Reviews Ltd.
Document Type: Report
Length: 3,384 words
Lexile Measure: 1410L

Document controls

Main content

Article Preview :

Author(s): Tamara MH Gall 1 , Adam E Frampton [*] 4 , Jonathan Krell 2 , Leandro Castellano 2 , Justin Stebbing 3 , Long R Jiao 1



biomarker; colorectal cancer; diagnostic; miR-21; miRNA

Many studies have investigated the potential use of miRNAs in cancer detection and also prognostication. This is an exciting area for biomarker discovery, and there is no doubt that miRNA profiles are useful in differentiating malignant tissue and biofluids from non-neoplastic controls. Finding specific circulating blood miRNAs may allow for easy tumor-specific detection or disease stratification. Kanaan et al . found that tumor-derived miR-21 was significantly upregulated in peripheral blood samples of patients with colorectal cancer (CRC) [1] . However, it seems unlikely that an individual miRNA, such as miR-21, will become a useful blood-based biomarker for diagnosing CRC, as it is also dysregulated in many other diseases [2] . Perhaps, plasma miR-21 levels may be able to differentiate adenomas from CRC, enable the surveillance of patients with intermediate/high-risk polyps or further classify CRC patients with regards to stage, survival or response to chemotherapeutics [3] . From a diagnostic perspective, a specific blood-based miRNA signature may be more sensitive for identifying a specific pathology. Ideally, blood-based miRNA profiles should be established for age-matched healthy individuals, patients with early lesions (i.e., colonic adenomas) and patients with various stages of CRC to make accurate future clinical assessments [3] .

Summary of methods & results


Kanaan et al . performed their study in four phases [1] . First, they identified dysregulated miRNAs in CRCs. Colorectal tumor and adjacent non-neoplastic samples were collected from 30 patients (the 'training' set). It is not stated if these frozen surgical samples underwent any dissection prior to RNA extraction. The samples were then analyzed by reverse transcription-quantitative PCR using a TaqMan® (Invitrogen, CA, USA) array card (measuring 380 miRNAs). The most dysregulated miRNAs were next validated in a further 16 CRC patients (the 'test' set). In the third stage, RNA was extracted from blood from the same initial 30 CRC patients and 30 healthy controls. The authors stated that their modified TRIzol® (Invitrogen) protocol yielded "pure and high plasma miRNA levels"; however, this method was not detailed. The most differentially expressed miRNAs in CRC tissue were subsequently measured in the blood samples. Finally, these findings were validated using blood from 20 different CRC patients and matched healthy individuals. Small nuclear RNA U6 was used for normalization. It is not stated whether the plasma samples were collected pre-, intra- or post-operatively.


In the tissue 'training set', ten miRNAs were significantly upregulated, including miR-21, and nine were downregulated in CRC compared with normal adjacent tissue (p < 0.05 and false discovery rate of 10%). The four most dysregulated miRNAs (upregulated: miR-31 and miR-135b; downregulated: miR-1 and miR-133a) were then validated in a 'test set' of patients. Expression of these four miRNAs accurately distinguished CRCs from non-neoplastic tissue with 100% sensitivity and 85% specificity (area under the curve [AUC]: 0.932). This four-miRNA classifier as well as the third most upregulated and downregulated miRNAs in CRC tissues...

Source Citation

Source Citation   

Gale Document Number: GALE|A321864207