基于microRNA表达的肿瘤起源分类

, 刘华,郭诗翔,2, 鞠景芳

(1.Department of Pathology, Stony Brook University, School of Medicine, Stony Brook, New York, 11794,USA; 2.陆军军医大学第一附属医院肝胆外科，重庆 400038)

·述评·

基于microRNA表达的肿瘤起源分类

AndrewFesler1, 刘华1,郭诗翔1,2, 鞠景芳1

(1.DepartmentofPathology,StonyBrookUniversity,SchoolofMedicine,StonyBrook,NewYork, 11794,USA; 2.陆军军医大学第一附属医院肝胆外科，重庆 400038)

由于采用标准诊断方法难以识别癌变的原发部位，诊断新发的癌症中约有3%～5%来源于原发部位不明的肿瘤。MicroRNAs(miRNAs)近来被证实能够协助病理学家提高对原发部位不明肿瘤的诊断准确性。本文将着重讨论基于肿瘤诊断的miRNA最新研究进展，及该领域的未来发展方向。

微小RNA；诊断；组织来源不明肿瘤

1 Introduction

Accurate diagnosis is crucial to successful cancer treatment. Although most cancer patients present with a primary tumor, a small percentage (3%-5%) of cancer patients with metastatic disease have an unknown primary origin. In these cases, standard diagnostic approaches involving physical examinations, imaging, and pathological evaluations fail to identify a primary tissue of origin for an identified metastatic cancer. Without proper diagnosis, these patients often have a poor prognosis, as broad spectrum chemotherapy must be employed rather than targeted therapies, shown to be effective against specific cancers[1-3]. As a result, it is essential that new approaches are developed to enhance our ability to identify the tissue of origin and genetic finger prints as diagnostic biomarkers will be incredibly useful in helping to address this challenge.

Tremendous amounts of effort have been devoted to finding diagnostic biomarkers for tumors with unknown primary origin. miRNA based diagnostic biomarkers have emerged as promising candidates for assisting pathologist in identifying tissue of origin of metastatic cancers with unknown primary origin[4]. miRNAs are a class of non-coding small RNAs that post-transcriptionally, and translationally regulate mRNAs with important roles in development, differentiation, apoptosis, autophagy, and tumorigenesis[5-7]. Systematic evaluation of miRNA stability by our group has shown that miRNA are stable informalin fixed paraffin embedded (FFPE) specimens, making them excellent biomarkercandidates[8]. Additionally, miRNAs exhibit differential expression in different types of cancer[9]. Thus miRNA expression profiling has great potential to help in identifying tissue of origin for CUP patients.

2 miRNA expression profiling for identification of tissue of origin in CUP patients

In the past decade, there has been much interest in using miRNA expression profiling to help in identifying tissue of origin in CUP patients. Studies by Rosenfeld et al. identified a set of 48 miRNA based classifiers to determine tumor tissue origins for 22 tumor tissues using miRNA expression measured by microarray[4]. They have developed a miRNA based decision tree to guide the decision process for tumor tissue of origin determination. When used in combination with the K-nearest-neighbors (KNN) classification algorithm, in an independent blinded test-set, they found 86% of the cases were accurately identified. Among metastatic samples, the sensitivity was 85% for high-confidence classifications. This work demonstrated that primary tumor samples can be used to augment training sets if considerations are made for certain markers, which is important as availability of metastatic tissue is limited. Many of the tissue types were classified with high confidence except bladder cancer due to the small number of clinical samples used for the profiling studies[4].With the aim of developing a standardized test for identifying tissue of origin using this approach, the microarray was replaced with qRT-PCR for analysis of miRNA expression. Even with the switch to qRT-PCR expression analysis, tissue of origin was identified with relatively high accuracy of prediction based on a large set (204) of archival FFPE tumor samples[10]. While these initial studies used samples that had known primary origin, in order to confirm the ability of this approach to accurately identify the tissue of origin, another study used the qRT-PCR based approach to investigate tumor samples from CUP patients. On a set of 57 CUP patients with brain metastasis, the tissue of origin prediction based on this approach matched the diagnosis based on available clinicopathological data in 80% of the samples[11]. In a study that built upon on the original Rosenfeld work, again using a binary decision tree and KNN classification, with the expression profile of 64 miRNAs and 42 tumor types,primary tissue of origin was identified with 85% accuracy in a set of 509 samples.In this study samples from CUP patients were also utilized and 88% agreement with clinicopathological based tissue of origin identification was found. In cases of CUP, this type of approach can be used to supplement, other clinicopathological data, to help confirm tissue of origin. A study looking at the correlation between tissue of origin based on clinicopathological data and miRNA expression profiling found the strongest agreement (92%) between final clinical diagnosis, and prediction based on miRNA expression. The agreement with the initial clinical presentation diagnosis was 70%. This shows that miRNA profiling can help to correctly identify the primary tissue of origin, which can help to guide therapeutic intervention[12]. Other groups have also shown miRNA expression profiles to be effective at identifying tissue of origin. One recent study demonstrated that miRNA based classifiers provided 88% accuracy foridentifying primary tumor sites based on metastatic tumor miRNA expression data from over 200 FFPE clinical specimens representing 15 different histologies[13]. Another study demonstrated 86% accuracy for identification of primary origin for metastatic cases using the expression of 47 miRNAs[14]. A number of the miRNA biomarkers used in these studies to distinguish tissue of origin, also have functional and prognostic significance in cancer[15-20].

These studies are largely retrospective in nature, and while these retrospective studies are essential to developing the approach and confirming its accuracy prospective studies are need to demonstrate its utility. A prospective study based on miRNA expression profiling demonstrated a close correlation (84% agreement) with clinical pathological diagnosis of tumors with unknown tissue origin[21]. This effort is highly significant in that it clearly demonstrates the clinical impact of aiding accurate diagnosis for planning proper treatment options for these patients. Tissue of origin predictions provided by miRNA profiling may be especially helpful when clinicopathological information suggests differential diagnosis.

3 miRNA expression and CUP biology

Beyond the usefulness of miRNA expression profiling for identifying tissue of origin for CUP, there is also interest in using miRNA to understand the biology of CUP. Pentheroudakis et al.investigated miRNA expression in favorable prognosis subgroups of CUP, and found no significant expression differences between CUP metastatic tumors and metastatic tumors with known primary origin[22]. This suggests that there is not a universal miRNA signature for CUP. Another study looked at differences in expression of miRNA in CUP tumors that are EMT positive vs. those that are EMT negative. This work found no statistically significant difference but did identify some miRNA with differential expression[23]. Clearly more work needs to be done to understand the biology of CUP tumors and the role miRNA may play in their development.

4 Summary and future perspective

miRNA profiling has shown great promise for helping to identify tissue of origin for CUP patients. With additional efforts, approaches based on miRNA expression profiling of CUP may be further refined to greatly enhance our ability to identify tissue of origin and accurately diagnosis these patients.Beyond tumor tissue based diagnosis, recent studies have demonstrated promising potential of liquid (e.g. serum, plasma, urine) sample miRNA based diagnosis[24]. It may be possible to identify metastatic cancer tissue origin based on miRNA expression analysis on circulating tumor cells or circulating microRNAs. Such tests will be non-invasive and may help to guide treatment decisions if tissue of primary origin can be identified. The application of such an approach will require multi- and inter-disciplinary team work and multi-center studies with large patient cohorts for discovery and validation. Hopefully miRNA expression profiling will help to improve outcomes for CUP patients.

Acknowledgements:This study was supported by National Institute of Health/National Cancer Institute R01CA15501904 (J. Ju), R01CA19709801 (J. Ju).

[1]Greco FA. Cancer of unknown primary site: evolving understanding and management of patients[J].Clin Adv Hematol Oncol, 2012,10(8):518-524.

[2]Varadhachary GR, Raber MN.Cancer of unknown primary site[J].N Engl J Med, 2014,371(8):757-765.

[3]Pavlidis N, Pentheroudakis G.Cancer of unknown primary site[J].Lancet, 2012,379(9824):1428-1435.

[4]Rosenfeld N, Aharonov R, Meiri E,etal. MicroRNAs accurately identify cancer tissue origin[J].Nat Biotechnol, 2008,26(4):462-469.

[5]Calin GA, Dumitru CD, Shimizu M,etal. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia[J].Proc Natl Acad Sci USA, 2002,99(24):15524-15529.

[6]Zhai H, Fesler A, Ju J. MicroRNA: a third dimension in autophagy[J].Cell Cycle, 2013,12(2):246-250.

[7]Karaayvaz M, Zhai H, Ju J.miR-129 promotes apoptosis and enhances chemosensitivity to 5-fluorouracil in colorectal cancer[J].Cell Death Dis, 2013,4:e659.

[8]Xi Y, Nakajima G, Gavin E, Systematic analysis of microRNA expression of RNA extracted from fresh frozen and formalin-fixed paraffin-embedded samples[J].RNA, 2007,13(10):1668-1674.

[9]Di Leva G, Croce CM.MiRNA profiling of cancer[J].Curr Opin Genet Dev, 2013,23(1):3-11.

[10]Rosenwald S, Gilad S, Benjamin S,etal. Validation of a microRNA-based qRT-PCR test for accurate identification of tumor tissue origin[J].Mod Pathol, 2010,23(6):814-823.

[11]Mueller WC, Spector Y, Edmonston TB,etal.Accurate classification of metastatic brain tumors using a novel microRNA-based test[J].Oncologist, 2011,16(2):165-174.

[12]Pentheroudakis G, Pavlidis N, Fountzilas G,etal. Novel microRNA-based assay demonstrates 92% agreement with diagnosis based on clinicopathologic and management data in a cohort of patients with carcinoma of unknown primary[J].Mol Cancer, 2013,12:57.

[13]Sφkilde R, Vincent M, Mφller AK,etal.Efficient identification of miRNAs for classification of tumor origin[J].J Mol Diagn, 2014,16(1):106-115.

[14]Ferracin M, Pedriali M, Veronese A,etal. MicroRNA profiling for the identification of cancers with unknown primary tissue-of-origin[J].J Pathol, 2011,225(1):43-53.

[15]Song B, Wang Y, Kudo K,etal.miR-192 Regulates dihydrofolate reductase and cellular proliferation through the p53-microRNA circuit[J].Clin Cancer Res, 2008,14(24):8080-8086.

[16]Nakajima G, Hayashi K, Xi Y,etal. Non-coding microRNAs hsa-let-7g and hsa-miR-181b are associated with chemoresponse to S-1 in colon cancer[J].Cancer Genomics Proteomics, 2006,3(5):317-324.

[17]Song B, Wang Y, Titmus MA,etal. Molecular mechanism of chemoresistance by miR-215 in osteosarcoma and colon cancer cells[J].Mol Cancer, 2010,9:96.

[18]Song B, Wang Y, Xi Y,etal. Mechanism of chemoresistance mediated by miR-140 in human osteosarcoma and colon cancer cells[J].Oncogene, 2009,28(46):4065-4074.

[19]Braun CJ, Zhang X, Savelyeva I,etal. p53-Responsive micrornas 192 and 215 are capable of inducing cell cycle arrest[J].Cancer Res, 2008,68(24):10094-10104.

[20]Cimmino A, Calin GA, Fabbri M,etal. miR-15 and miR-16 induce apoptosis by targeting BCL2[J].Proc Natl Acad Sci USA, 2005,102(39):13944-13949.

[21]Varadhachary GR, Spector Y, Abbruzzese JL,etal. Prospective gene signature study using microRNA to identify the tissue of origin in patients with carcinoma of unknown primary[J].Clin Cancer Res, 2011, 17(12):4063-4070.

[22]Pentheroudakis G, Spector Y, Krikelis D,etal. Global microRNA profiling in favorable prognosis subgroups of cancer of unknown primary (CUP) demonstrates no significant expression differences with metastases of matched known primary tumors[J].Clin Exp Metastasis, 2013,30(4):431-439.

[23]Stoyianni A, Pentheroudakis G, Benjamin H,etal.Insights into the epithelial mesenchymal transition phenotype in cancer of unknown primary from a global microRNA profiling study[J].Clin Transl Oncol, 2014,16(8):725-731.

[24]Fesler A, Jiang J, Zhai H,etal.Circulating microRNA testing for the early diagnosis and follow-up of colorectal cancer patients[J].Mol Diagn Ther, 2014,18(3):303-308.

2017-09-30)

(本文编辑：许晓蒙)

MicroRNA expression based tumor origin classification

AndrewFesler1,LIUHua1,GUOShi-xiang1,2,JUJing-fang1

(1.DepartmentofPathology,StonyBrookUniversity,SchoolofMedicine,StonyBrook,NewYork, 11794USA; 2.InstituteofHepatopancreatobiliarySurgery,SouthwestHospital,TheFirstHospitalAffiliatedtoAMU,Chongqing400038,China)

Approximately 3 to 5% of newly diagnosed metastatic cancers are of unknown primary tissue origin due to difficulties identifying a primary tumor using standard diagnostic approaches.MicroRNAs (miRNAs) have recently been demonstrated to be able to assist pathologist with improved accuracy in diagnosing cancers of unknown primary origin (CUP). In this short commentary, we will highlight some of the recent advancements in miRNA based cancer diagnosis as well as some future directions for the field.

microRNA; diagnosis; tumor with unknown tissue origin

10.13602/j.cnki.jcls.2017.11.01

Andrew Fesler,1991年生，男，助理研究员，从事肿瘤研究工作。

鞠景芳，教授，博士，E-mail:jingfang.ju@stongbrookmedicine.edu。

R730.4

A