Comprehensive Evaluation Of Circulating MicroRNAs As Diagnostic Biomarkers For Colorectal Cancer

Background Colorectal cancer(CRC)is a commonly diagnosed cancer worldwide.Although population-based screening has been proved to effectively prevent CRC at early stage,approximately 60% of CRC patients are diagnosed at regional or distant stage with a dissatisfied 5-year survival rate ranging from 12.5% to 70.4%.Moreover,current tumor markers such as carcino-embryonic antigen(CEA)and CA125 are found to be useless alone in early detection of CRC,inevitably resulting in delayed diagnosis.Recent studies have provided numerous molecular biomarkers for disease detection and prognostic prediction in CRC patients.Exploring more efficact and accuracy diagnostic strategy is therefore urgently needed to accurately diagnose CRC and guide treatment.Previous work found tumor-derived mi RNAs can be released into circulation in the form of exosome,microvesicles or through combining with RNA binding proteins and lipoproteins,suggesting non-invasive detecting circulating mi RNAs may be a novel diagnostic approach for CRC patients.Despite the number of CRC-specific mi RNAs biomarkers remain keeping in growth,the clinical applicability of those reported mi RNAs are now in a dilemma for the accumulating inconsistencies among different institutes.[16,17]Meanwhile,the issue that how the outcomes of previous mi RNAs biomarkers studies should be integrated into clinical setting is still unexplored.Thus,this study aimed to broadly elucidate the replicability of the identified mi RNAs biomarkers of CRC and to develop a feasible testing strategy for future clinical practice based on a highly standardized platform within an independent Chinese cohort.Materials and MethodsThe study generally consisted of four parts: a systematic literature review for selecting candidate mi RNAs,a training phase for screening candidate mi RNAs and constructing diagnostic models,a validation phase for confirming optimal mi RNAs/diagnostic models,and a supplement phase for investigating the expression of optimal mi RNAs in precancerous lesions.We initially performed a systematic literature review to confirm candidate mi RNAs for research.A three phase laboratory study then was performed.First,we examined the expression levels of selected mi RNAs in a training set of 120 plasma samples(CRC vs HC=60:60)by quantitative real-time PCR method(q RT-PCR).Subsequently,binary logistic regression analysis was adopted to develop a blood-based diagnostic model.Another cohort of 160 plasma samples(CRC vs HC=80:80)were used as a validation set.Additionally,we enrolled 17 patients with advanced adenomas(AA),and 25 patients with inflammatory bowel disease(IBD)to investigate the expression patterns of the dysregulated mi RNAs in precancerous diseases.Results We firstly selected 30 candidate mi RNAs(mi R-15 b,mi R-17-3p,mi R-17,mi R-18a-3p,mi R-18 a,mi R-19 a,mi R-19 b,mi R-21,mi R-24,mi R-26 b,mi R-29 a,mi R-34 a,mi R-92 a,mi R-125 b,mi R-139-3p,mi R-145,mi R-150,mi R-183,mi R-194,mi R-221,mi R-224,mi R-320 a,mi R-331-3p,mi R-372,mi R-378,mi R-423,mi R-431,mi R-592,mi R-601 and mi R-760)from a systematic literature review.Then,the expression of 30 selected mi RNAs was examined using q RT-PCR in a training cohort of 120 plasma samples(CRC vs health control(HC)=60:60),and confirmed in a validation cohort of 160 plasma samples(CRC vs HC=80:80),and finally investigated in a supplementary cohort of 42 plasma samples(17 for advanced adenomas(AA)and 25 for ulcerative colitis(UC)).As a result,we finally determined 9 dysregulated circulating mi RNAs(mi R-15 b,mi R-17,mi R-19 a,mi R-21,mi R-24,mi R-26 b,mi R-145,mi R-592 and mi R-760)from the 30 candidates,among which mi R-21 and mi R-26 b have the best diagnostic performance.Meanwhile,we proposed a novel model integrating 4 mi RNAs(mi R-26 b,mi R-24,mi R-592 and mi R-760)and serum CEA for noninvasive blood-based diagnosis,which has a significantly better diagnostic performance than 4 mi RNAs or CEA alone or any single mi RNA.Finally,we found circulating mi R-17 and mi R-145 were dysregulated in AA patients compared with HC,suggesting their potential role as early detection of precancerous lesions.Conclusion In summary,our study comprehensively validated the diagnostic value of currently reported circulating micro RNAs in CRC and proposed that combined detection of multiple circulating mi RNAs and serum CEA is a promising noninvasive blood-based diagnostic strategy.