脑脊液检测

LABORATORY INVESTIGATIONMicroRNAs in cerebrospinal fluid as biomarker for disease course monitoring in primary central nervous system lymphomaAlexander Baraniskin •Jan Kuhnhenn •Uwe Schlegel •Wolf Schmiegel •Stephan Hahn •Roland SchroersReceived:6March 2012/Accepted:29May 2012/Published online:23June 2012ÓSpringer Science+Business Media,LLC.2012Abstract Diagnosis of primary lymphomas of the central nervous system (PCNSL)largely depends on histopathol-ogy of tumor biopsies.Recently,we identified miRNAs detected in the CSF of PCNSL patients as novel non-invasive biomarkers for this disease.In combined analyses of miR -21,miR -19b ,and miR -92CSF levels,it was pos-sible to differentiate PCNSL from other neurological dis-orders.In the current study,we first confirmed our previous findings in an enlarged PCNSL cohort (n =39;sensitivity 97.4%).Also,we sought to establish the potential role of CSF miRNAs as biomarkers for disease course monitoring.In sequential miRNA measurements in CSF derived from nine patients with different disease courses,an intriguing correlation of miRNA levels and PCNSL status during treatment and/or disease follow-up was demonstrated.Finally,we demonstrated that miRNA levels in serum of PCNSL patients (n =14)were not elevated as comparedto controls.In summary,this study provides the first evi-dence that CSF miRNAs have the potential as biomarkers for treatment monitoring and disease follow-up of patients with PCNSL.Keywords Primary central nervous system lymphoma ÁCerebrospinal fluid (CSF)ÁSerum ÁMicroRNA (miRNA)ÁDisease courseIntroductionPrimary lymphomas of the central nervous system (PCNSL)represent a subcategory of extranodal non-Hodgkin lymphomas (NHL)confined to the CNS [1].Because PCNSL represent highly aggressive brain tumors,early diagnosis is essential for successful treatment and potential improvement of prognosis [2].Stereotactic needle biopsy including histopathology remains the standard diagnostic procedure for patients with suspected PCNSL [3].However,definitive histopathological diagnosis cannot always be achieved,especially in patients treated with corticosteroids.Identification of biomarkers in the cere-brospinal fluid (CSF)in order to accelerate the non-inva-sive diagnosis of PCNSL represents an attractive research goal.To date,the diagnosis of recurrent or progressive PCNSL mostly relies on neuroimaging.Hence,identifica-tion of novel biomarkers such as microRNAs for disease course monitoring is also an attractive research goal.Recently,we have demonstrated that microRNAs detected in the CSF by real-time quantitative polymerase chain reaction can serve as non-invasive biomarkers for PCNSL [4].MicroRNAs are an abundant class of small non-protein-coding RNA molecules that function as nega-tive gene regulators by silencing gene expression at aElectronic supplementary material The online version of this article (doi:10.1007/s11060-012-0908-2)contains supplementary material,which is available to authorized users.A.Baraniskin ÁW.Schmiegel ÁR.Schroers (&)Department of Medicine,Hematology and Oncology,Ruhr-University of Bochum,Knappschaftskrankenhaus Bochum-Langendreer,In der Schornau 23/25,44892Bochum,Germany e-mail:Roland.Schroers@rub.deA.Baraniskin ÁS.HahnCenter of Clinical Research,Ruhr-University of Bochum,Bochum,GermanyJ.Kuhnhenn ÁU.SchlegelDepartment of Neurology,Ruhr-University of Bochum,Bochum,GermanyJ Neurooncol (2012)109:239–244DOI 10.1007/s11060-012-0908-2post-transcriptional level by means of pleiotropic sup-pression of sequence-complementary mRNA targets[5].In our previous study,combined expression analyses of miR-21,miR-19b,and miR-92in CSF revealed that CSF levels of miRNA could differentiate,with high specificity (96.7%)and sensitivity(95.7%),patients with PCNSL from other neurologic disorders[4].In the current study,we were able to confirm our pre-vious results in an expanded cohort of PCNSL patients.In addition,we analyzed miRNA levels in the CSF of indi-vidual PCNSL patients at different time points following diagnosis,questioning whether miRNAs could serve as markers of disease course.Furthermore,we investigated whether miR-21,miR-19b,and miR-92can also be detected in serum of PCNSL patients as compared to control subjects.Materials and methodsPatient characteristics,CSF samples,serum samples, microRNA quantificationCSF samples from39immunocompetent patients,in which a histopathological diagnosis of diffuse large B cell type lymphoma(DLBCL)had been established,were included in this study.CSF samples were collected before chemo-therapy by diagnostic lumbar puncture after written informed consent.The University of Bochum Ethical Committee had approved CSF and peripheral blood sample collections.The characteristics of the control group (n=37)have been described in our previous publication [4,6].Consecutive CSF samples were collected by diag-nostic lumbar puncture or puncture of Ommaya reservoirs from individual PCNSL patients during treatment and during disease follow-up.Serum samples were obtained at the time point of CSF sample collections.Serum samples were centrifuged(5009g,10min,room temperature)to remove cells and debris and were stored at-80°C until further processing.Preparation of CSF and serum samples including RNA extraction,reverse-transcription,and microRNA quantification by real-time polymerase reaction were performed as has recently been described in detail[4]. StatisticsStatistical analyses were performed with SPSS(v.20;SPSS) and GraphPad Prism(v.5.0).Group-wise comparisons of distributions of clinical and biologic data were performed, applying2-tailed Mann–Whitney U tests and Kruskal–Wallis tests with Dunn multiple comparisons.Results were considered statistically significant at P values\0.05.ResultsCirculating microRNAs in CSF as biomarkerfor detection of PCNSLTo confirm and expand the results of our recent study unraveling miR-21,miR-19,and miR-92CSF levels for PCNSL diagnosis,we included16patients with newly diagnosed PCNSL into our original cohort of23patients [4].The expanded PCNSL cohort(n=39)consisted of20 male and19female patients;age was between42and 87years(mean age,66years).At the time of CSF col-lection,the disease was newly diagnosed in35of39 PCNSL patients.Altogether,21of39patients(54%)were treated with corticosteroids at the time of CSF sample collection.Making use of miRNA quantification by real-time polymerase chain reactions,the levels of the four miRNAs miR-24,miR-21,miR-19b,and miR-92were quantified in each CSF sample as recently reported[4]. Applying normalization of the amount of target miRNA relative to the amount of miR-24and considering REL above the cut-offs for miR-21and for one of each miR-19b or miR-92,respectively[4],all new patients were correctly classified as PCNSL,independent of prior corticosteroid treatment(Supplementary Table S1).In summary,38of39 PCNSL patients were properly identified.Thus,the sensi-tivity of combined CSF analyses of miR-21,miR-19b,and miR-92was97.4%in our enlarged PCNSL cohort. Circulating miRNAs in CSF as biomarkers of disease course of PCNSLLongitudinal studies of miRNA levels in CSF of nine patients with PCNSL were performed.For this purpose, sequential CSF samples were collected at diagnosis und at different time points during disease course by means of lumbar puncture prior to chemotherapy and by CSF sample collection from Ommaya reservoirs which are routinely used for intra-ventricular chemotherapy in our standard treatment protocol[7].Dependent on outcome,the patients were classified into three groups:(1)continuing complete remission of PCNSL(n=5);(2)recurrent PCNSL after initial remission(n=2);and(3)progressive disease despite therapy(n=2;Table1).As previously described [4],samples with relative expression levels(REL)above the cut-off for miR-21and above cut-offs for one of each miR-19b or miR-92,respectively,were classified as ‘‘miRNA status positive’’(Table1;Fig.1).As cut-offs,the following REL were used:8.0REL for miR-21,1.4REL for miR-19b,and2.5REL for miR-92[4].Longitudinal REL data of each indicative miRNA in the CSF of all nine patients correlated well with the clinical courses(Fig.1; Table1).Five PCNSL patients with continuous completeremission showed a marked decrease of miRNA expression levels and turned from‘‘positive’’to‘‘negative’’miRNA status in the CSF(Table1),as exemplified for patient1in Fig.1a.In two patients(6and7,Table1)transient responses to chemotherapy followed by progressive PCNSL were demonstrated in consecutive MRIs.In accordance,the CSF miRNA levels initially decreased and subsequently increased during disease progression as determined by qRT-PCR(Fig.1b).Of note,the microRNA status in the CSF of patient6turned back from‘‘negative’’to‘‘positive’’and accordingly indicated PCNSL relapse 72days before relapse was detected by MRI.Finally, markedly increasing levels of CSF miR-21,miR-19b,and miR-92were observed in one patient with primary pro-gressive disease(patient8,Table1)and in another patient relapsing after complete remission for12months(patient 9,Table1;Fig.1c).Interestingly,the CSF miRNA levels correlated to the tumor volume as measured in MRIs (Fig.1).These data support the diagnostic value of CSF miRNA levels for disease course monitoring.Importantly, in contrast to CSF miRNA,standard CSF parameters such as total cell count and protein concentration did not show consistent changes correlating with the PCNSL status during treatment and follow-up.Serum miRNAs in PCNSL patientsDetection of PCNSL related miRNAs in serum would facilitate the diagnostic procedure exploiting these novel biomarkers.Accordingly,we investigated whether the levels of PCNSL biomarkers miR-21,miR-19b,and miR-92 as detected in serum specimens differed between PCNSL patients(n=14)and control patients with various neuro-logic disorders(headache,epilepsy,syncope and stroke; n=8).Total RNA was isolated from frozen serum sam-ples and miRNA levels were measured in qRT-PCR asTable1PCNSL patients: clinical characteristics and relative miRNA expression levels in CSFL lumbar,O Ommaya reservoir Patient DiseasestageCSFsourceRELmiR-21RELmiR-19bRELmiR-92miRNAstatus PCNSL patients with persistent complete remission1Diagnosis L138.524.390.4Positive Therapy L 2.80.5 1.9NegativeCR L 1.70.6 2.0Negative 2Diagnosis L37.4 6.821.1Positive Therapy O7.50.8 3.7NegativeCR O 3.10.20.8Negative 3Diagnosis O11.1 1.47.1Positive Therapy O 2.00.4 2.0NegativeTherapy O0.40.10.4NegativeCR O0.80.20.6Negative 4Diagnosis L12.9 4.333.4Positive Therapy L 5.50.4 2.0Negative 5Diagnosis O21.7 3.917.0Positive Therapy O16.90.7 1.0NegativeCR O 2.90.10.5Negative PCNSL patients with transient remission6Diagnosis L10.9 2.1 6.8Positive Therapy O19.40.40.7NegativeTherapy O8.50.7 2.8PositiveRecurrence O9.0 2.17.7PositiveRecurrence L108.8 1.98.1Positive 7Diagnosis L10.1 4.59.6Positive Therapy L7.50.4 1.0NegativeRecurrence L10.9 1.10.02Negative PCNSL patients with recurrent disease8Diagnosis L 3.264.414.3Negative Progressive disease L276.3447.3129.3Positive 9CR L 6.2 1.6 4.9Negative Recurrence L18.09.811.8Positivepreviously reported [4].In these analyses we did not find significant differences in expression levels of miR -21,miR -19b,and miR -92in serum as demonstrated in Table 2.DiscussionRecently,we have demonstrated that combined qRT -PCR expression analyses of miR -21,miR -19b ,and miR -92in the CSF can serve as non-invasive biomarker for PCNSL [4].These three microRNAs were initially chosen in a candi-date approach based on our prior knowledge about their expression profiles in diffuse large B-cell lymphoma and primary CNS lymphoma tissue specimens.Moreover,important pleiotropic roles of our candidate microRNAs miR -21,miR -19b ,and miR -92in tumorigenesis and lym-phomagenesis are well established.Indeed,miR -21is a unique miRNA in that it is overexpressed in most tumor types analyzed thus far [8,9].Furthermore,higher levels of miR -21are associated with worse prognosis of DLBCL,and miR -21acts as a key oncogene in B-lymphomagenesis in vivo [8,10].Both miR -19b and miR -92are members of the poly-cistronic microRNA-17-92cluster and are over-expressed in DLBCL [9].In addition,the microRNA-17-92cluster promotes lymphomagenesis [11].Clustering of microRNA genes is an essential element of genomic organization.About 40%of miRNAs were found as part of a poly-citronic cluster [12].Although miR -19b and miR -92are adjacently located within the same primary microRNA transcript,different concentrations of both mi-croRNAs and different microRNA ratios are described in the literature [13]and were detected in our study (Table S1).These findings can be explained by various post-transcriptional mechanisms regulating microRNA biogen-esis and activity including the RNA-binding protein and Drosha processing [14].In addition,binding of some mi-croRNAs to proteins such as Argonaute2and their incor-poration into microvesicles protect microRNAs from degradation in body fluids [15,16].Therefore,even though miR -19b and miR -92are parts of the same poly-citronic cluster and are located on the same primary miRNA tran-script,they are characterized by different profiles in tissues and in body fluids.To further substantiate the results of our previous study [4]identifying miR -21,miR -19b ,and miR -92CSF expression for the purpose of PCNSL detection,we expanded the original cohort with 16patients.In the enlarged PCNSL cohort,the high sensitivity of 97.4%was demonstrated,viz.38of 39patients with PCNSL were classified correctly.These data confirm our previousmiRNA status tumor volume cell count positive 3500 42 53 negative not apparent4 43 negative not apparent2 45miRNA status tumor volume cell count positive 44051 2 negative 6000 1 positive 6000 3 positive 1440 (new lesion)3 positive 20000 6 miRNA status tumor volume negative not apparent 5 R ETable 2Serum levels of miRNAs in patients with newly diagnosed PCNSL versus control patientsPCNSL patients (n =14)Control patients (n =8)P valueCt aSD b Ct SD miR -2122.600.1722.630.580.58c miR -19b21.840.1422.370.530.43c miR -9220.740.1621.060.530.74c miR -2422.570.2223.180.540.48c a Data are means of CT values (groupwise)b Standard deviationcThe P value is for comparison of miRNA expression among PCNSL patients and control patients and was calculated using the Mann–Whitney U testfindings and further support the potential of miRNA quantification in the CSF as a non-invasive test for PCNSL.PCNSL patients frequently suffer from lymphoma relapse or disease progression.Here,diagnosis is primarily based on neuroimaging.Considering previous reports on miRNA detection in plasma during the course of sys-temic non-Hodgkin’s lymphomas[17],we addressed CSF miRNA levels over time of individual PCNSL patients as potential biomarkers for the purpose of disease follow-up. In our pilot study,longitudinal determinations of miRNA levels in CSF of nine patients with PCNSL were per-formed.Serial miRNA measurements of all nine patients agreed with tumor status based on objective criteria based on neuroimaging and correlated to the tumor volume as measured in MRIs.Of note,contrary to CSF miRNA, currently established standard CSF parameters such as total cell count and protein concentration did not show any changes correlating with the PCNSL disease course.A crucial feature of an ideal PCNSL biomarker is that its differential expression indicates the course of disease and provides insight into disease status with greater sensitivity than other diagnostic methods.Measurements of individual CSF miRNA expression profiles within prospective clinical studies including larger patient samples may ultimately facilitate early,noninvasive diagnosis,risk stratification, and determination of appropriate therapeutic interventions in patients with indeterminatefindings in neuroimaging. Although the number of patients studied here is relatively small,our study provides the rationale for future investi-gations of miRNAs in CSF as follow-up biomarkers of PCNSL patients in a larger cohort.Accumulating reports suggest that circulating miRNAs are detectable in peripheral blood and have the potential to be utilized as biomarkers for diverse malignancies[18,19]. Previously,Lawrie et al.demonstrated that miR-155,miR-210,and miR-21were elevated in the serum of de novo DLBCL patients,including CNS involvement in one case. Furthermore,miR-21expression was associated with relapse free survival[20].Ohyashiki et al.[17]reported about miR-92in plasma as a novel biomarker not only for diagnosis but also for monitoring DLBCL patients after chemotherapy.In the current study,we found no significant differences in expression levels of miR-21,miR-19b,and miR-92in serum between PCNSL patients and control patients.A possible reason for this conflicting result is the blood–brain barrier which separates CNS lymphomas from blood circulation[21].Thus,detection of miRNAs derived from PCNSL is apparently restricted to the CNS com-partment including the CSF.In conclusion,we validated our previously identified PCNSL-associated CSF miRNA profile in an extended patient cohort.We further demonstrated that CSF miRNAs have a potential as novel biomarkers not only for disease diagnosis but also for therapeutic and follow-up monitoring of individual PCNSL patients.Wefinally showed that serum miRNA levels in PCNSL patients are not helpful in diagnostic differentiation from controls with other neuro-logical disease.Acknowledgments This study was supported by a grant from the Ruhr-University of Bochum(FORUM),and by a grant(PURE)from the Ministry of Science,North Rhine-Westphalia,Germany.We thank Andriy Alekseyev for analyzing the MRI scans.Also,we thank S.Plambeck and G.Kersten for excellent technical assistance.Conflict of interest All authors declare that there is no potential conflict of interest.References1.Kluin PM,Deckert M,Ferry JA(2008)Primary diffuse largeB-cell lymphoma of the CNS.In:Swerdlow SH,Campo E,Harris NL,Jaffe ES,Pileri S,Stein H,Thiele J,Vardiman JW(eds) WHO classification of tumours of haematopoietic and lymphoid tissues.WHO,Lyon,pp240–2412.Deckert M,Engert A,Bruck W,Ferreri AJ,Finke J,Illerhaus G,Klapper W,Korfel A,Kuppers R,Maarouf M,Montesinos-Rongen M,Paulus W,Schlegel U,Lassmann H,Wiestler OD, Siebert R,Deangelis LM(2011)Modern concepts in the biology, diagnosis,differential diagnosis and treatment of primary central nervous system lymphoma.Leukemia25:1797–18073.Baraniskin A,Deckert M,Schulte-Altedorneburg G,Schlegel U,Schroers R(2011)Current strategies in the diagnosis of diffuse large B-cell lymphoma of the central nervous system.Br J Haematol156:421–4324.Baraniskin A,Kuhnhenn J,Schlegel U,Chan A,Deckert M,GoldR,Maghnouj A,Zollner H,Reinacher-Schick A,Schmiegel W, Hahn SA,Schroers R(2011)Identification of microRNAs in the cerebrospinalfluid as marker for primary diffuse large B-cell lymphoma of the central nervous system.Blood117:3140–3146 5.Ventura A,Jacks T(2009)MicroRNAs and cancer:short RNAsgo a long way.Cell136:586–5916.Baraniskin A,Kuhnhenn J,Schlegel U,Maghnouj A,Zollner H,Schmiegel W,Hahn S,Schroers R(2011)Identification of mi-croRNAs in the cerebrospinalfluid as biomarker for the diagnosis of glioma.Neuro Oncol14:29–337.Juergens A,Pels H,Rogowski S,Fliessbach K,Glasmacher A,Engert A,Reiser M,Diehl V,Vogt-Schaden M,Egerer G, Schackert G,Reichmann H,Kroschinsky F,Bode U,Herrlinger U,Linnebank M,Deckert M,Fimmers R,Schmidt-Wolf IG, Schlegel U(2010)Long-term survival with favorable cognitive outcome after chemotherapy in primary central nervous system lymphoma.Ann Neurol67:182–1898.Medina PP,Nolde M,Slack FJ(2010)OncomiR addiction in anin vivo model of microRNA-21-induced pre-B-cell lymphoma.Nature467:86–909.Robertus JL,Harms G,Blokzijl T,Booman M,de Jong D,vanImhoff G,Rosati S,Schuuring E,Kluin P,van den Berg A(2009) Specific expression of miR-17-5p and miR-127in testicular and central nervous system diffuse large B-cell lymphoma.Mod Pathol22:547–555wrie CH,Soneji S,Marafioti T,Cooper CD,Palazzo S,Pat-erson JC,Cattan H,Enver T,Mager R,Boultwood J,Wainscoat JS,Hatton CS(2007)MicroRNA expression distinguishes between germinal center B cell-like and activated B cell-likesubtypes of diffuse large B cell lymphoma.Int J Cancer 121:1156–116111.He L,Thomson JM,Hemann MT,Hernando-Monge E,Mu D,Goodson S,Powers S,Cordon-Cardo C,Lowe SW,Hannon GJ, Hammond SM(2005)A microRNA polycistron as a potential human oncogene.Nature435:828–83312.Altuvia Y,Landgraf P,Lithwick G,Elefant N,Pfeffer S,AravinA,Brownstein MJ,Tuschl T,Margalit H(2005)Clustering and conservation patterns of human microRNAs.Nucleic Acids Res 33:2697–270613.Thomson JM,Newman M,Parker JS,Morin-Kensicki EM,Wright T,Hammond SM(2006)Extensive post-transcriptional regulation of microRNAs and its implications for cancer.Genes Dev20:2202–220714.Chakraborty S,Mehtab S,Patwardhan A,Krishnan Y(2012)Pri-miR-17-92a transcript folds into a tertiary structure and autore-gulates its processing.RNA18:1014–102815.Arroyo JD,Chevillet JR,Kroh EM,Ruf IK,Pritchard CC,GibsonDF,Mitchell PS,Bennett CF,Pogosova-Agadjanyan EL, Stirewalt DL,Tait JF,Tewari M(2011)Argonaute2complexes carry a population of circulating microRNAs independent of vesicles in human plasma.Proc Nat Acad Sci USA108:5003–500816.Valadi H,Ekstrom K,Bossios A,Sjostrand M,Lee JJ,Lotvall JO(2007)Exosome-mediated transfer of mRNAs and microRNAs isa novel mechanism of genetic exchange between cells.Nat CellBiol9:654–65917.Ohyashiki K,Umezu T,Yoshizawa S,Ito Y,Ohyashiki M,Ka-washima H,Tanaka M,Kuroda M,Ohyashiki JH(2011)Clinical impact of down-regulated plasma miR-92a levels in non-Hodg-kin’s lymphoma.PLoS One6:e1640818.Hu Z,Chen X,Zhao Y,Tian T,Jin G,Shu Y,Chen Y,Xu L,ZenK,Zhang C,Shen H(2010)Serum microRNA signatures iden-tified in a genome-wide serum microRNA expression profiling predict survival of non-small-cell lung cancer.J Clin Oncol 28:1721–172619.Shen J,Todd NW,Zhang H,Yu L,Lingxiao X,Mei Y,GuarneraM,Liao J,Chou A,Lu CL,Jiang Z,Fang H,Katz RL,Jiang F (2011)Plasma microRNAs as potential biomarkers for non-small-cell lung b Invest91:579–587wrie CH,Gal S,Dunlop HM,Pushkaran B,Liggins AP,Pul-ford K,Banham AH,Pezzella F,Boultwood J,Wainscoat JS, Hatton CS,Harris AL(2008)Detection of elevated levels of tumour-associated microRNAs in serum of patients with diffuse large B-cell lymphoma.Br J Haematol141:672–67521.Teplyuk NM,Mollenhauer B,Gabriely G,Giese A,Kim E,Smolsky M,Kim RY,Saria MG,Pastorino S,Kesari S,Kri-chevsky AM(2012)MicroRNAs in cerebrospinalfluid identify glioblastoma and metastatic brain cancers and reflect disease activity.Neuro Oncol14:689–700。