Others

A variant of the LRP4 gene affects the risk of chronic lymphocytic leukaemia transformation to Richter syndrome

Description
A variant of the LRP4 gene affects the risk of chronic lymphocytic leukaemia transformation to Richter syndrome
Categories
Published
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A variant of the  LRP4  gene affects the risk of chroniclymphocytic leukaemia transformation to Richter syndrome Richter syndrome (RS) represents the transformation of chronic lymphocytic leukaemia (CLL) to aggressivelymphoma, most commonly diffuse large B-cell lymphoma(DLBCL) (Tsimberidou & Keating, 2005; Omoti & Omoti,2008; Rossi & Gaidano, 2009). Mechanisms and risk factors of CLL transformation to RS are known only in part (Tsimberi-dou & Keating, 2005; Omoti & Omoti, 2008; Rossi andGaidano, 2009). To date, predictors of RS transformationmainly consist of biological features related to the CLL clone,including usage of specific  IGHV   genes, expression of stereo- Silvia Rasi, 1 * Valeria Spina, 1 * AlessioBruscaggin, 1 Tiziana Vaisitti, 2 ClaudioTripodo, 3 Francesco Forconi, 4 LorenzoDe Paoli, 1 Marco Fangazio, 1 Elisa Sozzi, 4 Emanuele Cencini, 4 Luca Laurenti, 5 Roberto Marasca, 6 Carlo Visco, 7 Zijun Y.Xu-Monette, 7 Valter Gattei, 8 Ken H.Young, 7 Fabio Malavasi, 2 Silvia Deaglio, 2 Gianluca Gaidano 1 and Davide Rossi 1 1 Division of Haematology, Department of Clinical and Experimental Medicine & IRCAD, Amedeo Avogadro University of Eastern Piedmont, Novara,  2 Department of Genetics, Biology and Biochemistry, University of Turin, Turin, 3 Department of Human Pathology, University of Palermo, Palermo,  4 Division of Haematology,University of Siena and AOUS, Siena,  5 Division of Haematology, Catholic University of the Sacred Heart, Rome,  6  Division of Haematology,Department of Oncology and Haematology,University of Modena and Reggio Emilia, Modena, Italy,  7  Department of Pathology and Laboratory Medicine, University of WisconsinSchool of Medicine and Public Health, University of Wisconsin Paul P. Carbone ComprehensiveCancer Center, Madison, WI, USA, and   8 Clinical and Experimental Onco-Haematology Unit,Centro di Riferimento Oncologico, I.R.C.C.S., Aviano, Italy  Received 29 July 2010; accepted for publication1 October 2010Correspondence: Davide Rossi, MD, PhD,Division of Haematology, Department of Clinical and Experimental Medicine & BRMA,Amedeo Avogadro University of EasternPiedmont, Via Solaroli 17, 28100 Novara, Italy.E-mail: rossidav@med.unipmn.it*S.R. and V.S. contributed equally to this study. Summary Richter syndrome (RS) represents the transformation of chronic lymphocyticleukaemia (CLL) to aggressive lymphoma. Risk factors of CLL transformationto RS are only partly known. We explored the role of the host geneticbackground as a risk factor for RS occurrence. Forty-five single nucleotidepolimorphisms (SNPs) known to be relevant for CLL prognosis weregenotyped in a consecutive cohort of 331 CLL, of which 21 had transformedto RS. After correcting for multiple testing and adjusting for previously reported RS risk factors, the  LRP4  rs2306029 TT variant genotype was thesole SNP independently associated with a higher risk of RS transformation(Hazard Ratio: 4 Æ 17;  P   = 0 Æ 001;  q  = 0 Æ 047). The enrichment of   LRP4  TTgenotype in RS was confirmed in an independent series ( n  = 44) used forvalidation purposes. The LRP4 protein was expressed in CLL ( n  = 66).Bioinformatic analysis scored  LRP4  rs2306029 as a variant with possibledeleterious and damaging variant of LRP4.  LRP4  genotyping may help therecognition of patients with increased risk of RS at the time of CLL diagnosis. Keywords :  Richter syndrome, chronic lymphocytic leukaemia, diffuse largeB cell lymphoma, single nucleotide polimorphism, LRP4. research paper First published online 02 December 2010doi:10.1111/j.1365-2141.2010.08482.x  ª  2010 Blackwell Publishing Ltd,  British Journal of Haematology  ,  152,  284–294  typed B cell receptors, CD38 expression, cytogenetics, ortelomere length (Aydin  et al  , 2008; Rossi  et al  , 2008a, 2009a,b;Deambrogi  et al  , 2009).Increasing evidence points to the role of the geneticbackground of the host in the development and clinical courseof CLL. Several genotypic variants have been reported topredispose CLL development (Sellick   et al  , 2007; Di Bernardo et al  , 2008; Crowther-Swanepoel  et al  , 2010; Slager  et al  , 2010).Furthermore, single nucleotide polymorphisms (SNPs) haveemerged as relevant predictors of CLL outcome, both in thecontext of clinical trials and in consecutive CLL series (Nu¨ckel et al  , 2007; Gryshchenko  et al  , 2008; Rossi  et al  , 2008b; Sellick  et al  , 2008; Allan  et al  , 2010; Rasi  et al  , 2010).With respect to RS development, one SNP of the  CD38  gene(rs6449182) has been implicated in the acquisition of the RSclinical phenotype (Aydin  et al  , 2008). Indeed, RS appears tobe enriched in cases harbouring the  CD38  rs6449182 minor Gallele (Aydin  et al  , 2008). Overall, these data suggest that theinteraction between the genetic background of the host and thebiological characteristics of the neoplastic clone may affect theclinical course of CLL as well as the risk of RS transformation.This study aimed to further explore the role of the geneticbackground of the host as a risk factor for CLL transformationto RS. We report that a nonsynonimous SNP of the  LRP4  gene,a member of the low-density lipoprotein receptor gene family involved in the modulation of Wnt signalling, is a novel risk factor of RS transformation. Materials and methods Patients This study was based on a consecutive series of 331 CLLpatients, with typical sIg low/CD5 + /CD23 + phenotype andMatutes score >3 (Matutes  et al  , 1994; Hallek   et al  , 2008).Twenty-one (6 Æ 3%) of these patients had transformed toclonally related RS, as documented by the comparison of   IGH-V-D-J   rearrangements and HCDR3 sequences in CLL/RSpaired samples. Among CLL undergoing RS transformation,the median time from CLL diagnosis to RS was 29 Æ 1 months.For validation purposes, an independent cohort of 44 clonally related RS, as well as 7 cases of clonally unrelated RS, were alsocollected. RS diagnosis was based on histology of lymph nodeor extranodal tissue excisional biopsies. All RS cases wereclassified as DLBCL according to the World Health Organi-zation Classification of Tumours of the Haematopoietic andLymphoid Tissues (Swerdlow   et al  , 2008). Patients providedinformed consent in accordance with local IRB requirementsand Declaration of Helsinki. SNP genotyping  An educated guess approach was utilized for the selection of genes potentially relevant to RS transformation. SNPs wereselected according to the following criteria: (i) reportedassociation with CLL prognosis (Gryshchenko  et al  , 2008;Sellick   et al  , 2008; Allan  et al  , 2010; Rasi  et al  , 2010), (ii) minorallele frequency >5% in Caucasians. Accordingly, 45 SNPsfrom 45 genes were analysed (Table SI). SNP nomenclatureand nucleotide position were defined according to GenBank (http://www.ncbi.nlm.nih.gov. Accessed on 14 September2009). DNA samples for genotyping were extracted fromperipheral blood granulocytes collected at CLL diagnosis.Genotyping was performed by SNP minisequencing (ABIPrism SNaPshot Multiplex kit; Applied Biosystems, FosterCity, CA, USA), after validation of this approach by DNAdirect sequencing of each SNP in a pilot panel of cases ( n  = 60;data not shown). Quality control of genotyping was performedby replicate sample analysis.Deviation of SNP genotype distribution from Hardy–Weinberg equilibrium (HWE) was tested by chi-square testor Fisher’s exact test if appropriate. Seven SNPs were not inHWE (Table SI), and therefore were excluded from subse-quent analysis. Linkage disequilibrium (LD) among SNPs wascalculated as  r   2 values with the use of HAPLOVIEW (http://www.broad.mit.edu/haploview. Accessed on 14 September2009). None of the SNP was in strong LD ( r   2 ‡  0 Æ 8).  Analysis of LRP4 expression CLL cells were purified from peripheral blood mononuclearcells (PBMC) preparations by negative selection using anti-CD3, anti-CD16 and anti-CD14 monoclonal antibodies (mAbs;produced and purified in the Turin Laboratory) and magneticbeads (Invitrogen, Milan, Italy) separation. Purified CLL cellswere lysed in 1% Nonidet P-40 lysis buffer and whole cell lysates(25  l g/lane) were resolved by 10% sodium dodecyl sulphatepolyacrylamide gel electrophoresis and electrophoretically transferred to nitrocellulose membranes (Millipore, Billerica,MA, USA), using a semi-dry apparatus (Hoefer SemyPhor;Pharmacia Biotech, San Francisco, CA, USA). Blots wereblocked and incubated with a rabbit polyclonal antibody against LRP4 (Abgent, San Diego, CA, USA), followed by a goatanti-rabbit horseradish peroxidase-conjugated secondary reagent. An anti-actin mAb (Sigma, Milan, Italy) was used tocheck for homogeneous protein loading in each lane. Blots weredeveloped with an enhanced chemiluminiscence (ECL) detec-tion system (Perkin Elmer, Milan, Italy). Densitometric analysiswas performed using the public domain NIH I mage J program(version 1.36, available at rsb.info.nih.gov/nihimage/).For immunophenotypic studies, PBMC preparations werestained using a phycoerythrin (PE)- (for FACS studies) orAlexa568-labeled (for confocal studies) anti-CD19 mAb(produced in the University of Turin Laboratory), fixedand permeabilized using 0 Æ 1% saponin. Cells were thensaturated using pre-immune goat serum, before incubationwith the same rabbit polyclonal antibody against LRP4(Abgent) used for Western blot. Detection was done with anAlexa-488 conjugated goat anti-rabbit secondary antibody  Host Genetic Background and Risk of Richter Transformation ª  2010 Blackwell Publishing Ltd,  British Journal of Haematology  ,  152,  284–294  285  (Invitrogen). Cells were analysed using a FACSort (BDBiosciences, Milan, Italy), acquiring at least 5000 events persample using the Cell Q uest  software (Becton Dickinson,San Jose, California, USA). For confocal microscopy studies,labelled cells were left to adhere to polylysine-coated slides.Analysis was performed using an Olympus 1X71 FV300confocal microscope and the F luo V iew  software (Olympus,Milan, Italy).For immunohistochemical analysis of LRP4 expression, 30CLL bone marrow biopsies were collected from the archives of the Department of Human Pathology, University of Palermo,Italy. All bone marrow samples were collected at diagnosis,before any treatment was started. Four  l m-thick sections werecut from paraffin blocks and put onto slides. Followingdeparaffinization, sections underwent antigen retrieval by microwave oven treatment in citrate buffer pH 6 Æ 0 (Dako,Glostrup, Denmark). Subsequently, endogenous peroxidaseactivity was quenched by incubating slides with H 2 O 2  for10 min at room temperature. Sections were then incubatedwith a specific polyclonal antibody anti-human-LRP4 (SigmaLife Sciences, Sigma) overnight at 4  C. Binding of the primary antibody was revealed using a specific biotinylated secondary antibody, the streptavidin-biotin-peroxidase complex system(Novolink Max Polymer Detection System; Novocastra,Newcastle-upon-Tyne, UK), and 3-3 ¢ -diaminobenzidine aschromogenic substrate (brown signal). Slides were counter-stained with haematoxylin and evaluated under a LeicaDM2000 optical microscope (Leica Microsystems NusslochGmbH, Nussloch, Germany) by an expert hematopathologist.Microphotographs were collected using a Leica DFC320 digitalcamera (Leica). In silico analysis The consequence of   LRP4  rs2306029 on protein function wasanalysed  in silico  according to the PolyPhen (http://genet-ics.bwh.harvard.edu/pph. Accessed on 14 September 2009)and SIFT (http://blocks.fhcrc.org/sift/SIFT.html. Accessed on14 September 2009) algorithms. LRP4 protein sequenceinformation was retrieved from the National Center forBiotechnology Information (http://www.ncbi.nlm.nih.gov.Accessed on 12 April 2010) and the UniProtKB/Swiss-Prot(http://www.uniprot.org/. Accessed on 12 April 2010) data-bases. Alignment of amino acid sequences homologous tohuman LRP4 was computed using the M ega 4 (http://www.megasoftware.net) and C lustal W software (http://www.ebi.ac.uk/Tools/msa/clustalw2). Weblogo presentation of multi-ple sequence alignment around LRP4 Ser1554 was generatedusing the http://weblogo.berkeley.edu/ software (Accessed on12 April 2010). Statistical analysis The primary endpoint of the study was the cumulativeprobability of transformation measured from CLL diagnosis toRS transformation, death or last follow-up. The cumulativeprobability of transformation was calculated by Kaplan–Meiermethod (Kaplan & Meier, 1958). The association between SNPsand risk of RS transformation was actuarially assessed by univariate Cox analysis considering the minor allele as actingeither in a recessive or a in a dominant fashion. False discovery rate (FDR) was used to control for multiple statistical testing(Benjamini & Hochberg, 1995). Accordingly, along with the  P  value, also a  q  value was provided for each association analysis.The  q -value indicates the probability that a positive associationidentifiedbya P  -value<0 Æ 05mightbeafalsepositiveassociationdue to chance. Cox proportional hazard regression was used tobuild a multivariate model for RS transformation (Cox, 1972).Covariates included in multivariate analysis were  LRP4 rs2306029 genotype (CT/CC  versus  TT), HCDR3 status (nonstereotyped versus stereotyped),immunoglobulinvariableheavy chain ( IGHV  ) gene usage (non  IGHV4-39 versus IGHV4-39  ),and  CD38  rs6449182 genotype (CC  versus  CG/GG). In order toaccount for monotone likelihood, the Firth‘s bias correctionmethod was applied to Cox regression (Heinze & Schemper,2001). Unadjusted and adjusted hazard ratios for  LRP4 rs2306029 genotype were compared to estimate the proportionof the difference in the risk of RS transformation that wasattributable to covariates. Because the hazard ratio derived fromCox analysis measures magnitude of risk rather than a modelabilitytoaccuratelyclassifysubjects,Harrell c  -statisticswasusedto further evaluate the discriminatory value of the prognosticmodels in terms of RS development ( c  -statistics = 1 indicatesperfect discrimination, while  C   statistics = 0 Æ 5 equivalent tochance) (Harrell  et al  , 1996). To test whether the difference inthe discriminatory value of the prognostic models was signif-icant,weusedthercorrp.censfunctionintheDesignlibraryofR.Recursive-partitioning analysis for censored survival data wasperformed to identify the factors most influential for RStransformation, and to permit the classification of patients intorisk categories. Categorical variables were compared by chisquaretestandFisher’sexacttestwhenappropriate.Continuousvariables were compared by Mann–Whitney test. All statisticaltests were two-sided. Statistical significance was defined as P  -value <0 Æ 05. The analysis was performed with the StatisticalPackage for the Social Sciences ( spss ) software v.18.0 ( spss ,Chicago, IL, USA) and with  r  statistical package 2.8.1 (http://www.r-project.org. Accessed on 15 May 2010). Results The LRP4 rs2306029 genotype is an independent predictor of RS transformation In order to characterize the impact of the genetic backgroundof the host on the risk of RS transformation, 45 SNPs knownto associate with CLL progression and/or prognosis weregenotyped in the study cohort ( n  = 331) (Table SI). Theclinical characteristics at diagnosis of the CLL cohort areshown in Table I. After a median follow-up from diagnosis of  S. Rasi  et al  286  ª  2010 Blackwell Publishing Ltd,  British Journal of Haematology  ,  152,  284–294  75 months, 21/331 (6 Æ 3%) CLL transformed to RS, accountingfor a 5-year cumulative probability of transformation of 7 Æ 1%[95% confidence interval (CI) 3 Æ 8–10 Æ 4%].By Cox univariate analysis corrected for multiple compar-isons, the sole SNP that displayed a significant association withan increased risk of RS transformation was  LRP4  rs2306029( P   = 0 Æ 001;  q  = 0 Æ 047). CLL who carried the  LRP4  rs2306029TT variant genotype displayed a higher risk of transformation[Hazard Ratio (HR): 4 Æ 17; 95% CI: 1 Æ 75–9 Æ 92; transformed/  N  :12/79; 5-year risk: 13 Æ 5%; 95% CI: 5 Æ 3–21 Æ 7%] compared topatients carrying the CT/CC genotypes that contained the wildtype allele (transformed/  N  : 9/252; 5-year risk: 4 Æ 4%; 95% CI:1 Æ 5–6 Æ 9%) (Table II; Fig 1). Although the  CD38  rs6449182SNP was also significantly associated with RS transformationby Cox univariate analysis ( P   = 0 Æ 005), the probability that thisassociation was instead a false positive association due tochance was >10% ( q  = 0 Æ 117).The  LRP4  rs2306029 genotype appeared to specifically predict RS, whereas it had no impact on other CLL clinicalendpoints, including overall survival (HR: 1 Æ 24; 95% CI 0 Æ 78–1 Æ 97;  P   = 0 Æ 350) and time to first treatment (HR: 1 Æ 00; 95% CI0 Æ 70–1 Æ 43;  P   = 0 Æ 998).Other variables at CLL diagnosis associated with anincreased risk of RS were CD38 expression  ‡ 30%( P   = 0 Æ 011), unfavourable pattern of genetic lesions( P   = 0 Æ 004), IGHV homology   ‡ 98% ( P   = 0 Æ 002), stereotypedHCDR3 ( P   < 0 Æ 001), and usage of   IGHV4-39   ( P   < 0 Æ 001)(Table II). By statistical analysis, there was no evidence of anassociation between the above risk factors and the  LRP4 rs2306029 TT variant genotype ( P   > 0 Æ 050 in all instances),indicating that  LRP4  rs2306029 TT is not a surrogate of otherRS risk factors (Table III).By multivariate analysis, the  LRP4  rs2306029 TT genotypewas an independent predictor of RS transformation (HR: 3 Æ 21;95% CI: 1 Æ 22–8 Æ 45;  P   = 0 Æ 018), along with stereotyped HCDR3(HR: 4 Æ 45; 95% CI: 1 Æ 67–11 Æ 86;  P   = 0 Æ 003) and  IGHV4-39  usage (HR: 6 Æ 13; 95% CI: 1 Æ 95–19 Æ 19;  P   = 0 Æ 002) (Table II).The comparison between unadjusted and adjusted hazard ratiofor the  LRP4  rs2306029 TT genotype (Table II) revealed thatstereotyped HCDR3 and  IGHV4-39   usage could explain only 24% of the overall risk of RS transformation determined by the LRP4  rs2306029 TT genotype.  Model of RS prediction based on LRP4 rs2306029  genotype, HCDR3 status and IGHV4-39 usage Recursive-partitioning analysis was used to identify prognosticfactors with the most influential predictive significance for Table I.  Clinical features of the CLL cohort.Characteristics at diagnosisBiological features IGHV   homology   ‡ 98% 125/326 (38 Æ 3%)Stereotyped HCDR3 77/326 (23 Æ 6%) IGHV4-39   usage 9/326 (2 Æ 8%)FISH stratificationNormal FISH 89/320 (27 Æ 8%)del13q14 120/320 (37 Æ 5%)+12 55/320 (17 Æ 2%)del11q22-q23 22/320 (6 Æ 9%)del17p13 34/320 (10 Æ 6%) TP53  mutations 32/306 (10 Æ 5%)CD38  ‡ 30% 104/326 (31 Æ 9%)Clinical featuresAge (years) 68 (60–74)Male 175/331 (52 Æ 9%)Binet stageA 256/331 (77 Æ 3%)B 41/331 (12 Æ 4%)C 34/331 (10 Æ 3%)Nodal areas  ‡ 3 52/331 (15 Æ 7%)Largest node  ‡ 3 cm 34/331 (10 Æ 3%)Absolute lymphocyte count ( · 10 9 /l) 11 Æ 5 (6 Æ 4–20 Æ 0)Hb (g/l) 136 (128–148)Platelets ( · 10 9 /l) 193 (149–235)LDH (u/l) 360 (305–428)25th–75th percentiles are reported in parentheses for continuousvariables;  IGHV  , immunoglobulin heavy chain variable gene; HCDR3,immunoglobulin heavy chain complementarity determining region 3;FISH, fluorescence  in situ  hybridization. Table II.  Univariate and multivariate Cox analysis for cumulative probability of RS transformation.Univariate analysis Multivariate analysisHR LCI UCI  P   HR LCI UCI  P LRP4  rs2306029 TT 4 Æ 17 1 Æ 75 9 Æ 92 0 Æ 001 3 Æ 21 1 Æ 22 8 Æ 45 0 Æ 018 CD38  rs6449182 CG/GG 3 Æ 63 1 Æ 46 9 Æ 00 0 Æ 005CD38  ‡ 30% 3 Æ 26 1 Æ 31 8 Æ 13 0 Æ 011 TP53  deletion/mutation 3 Æ 81 1 Æ 53 9 Æ 51 0 Æ 004 IGHV   homology   ‡ 98% 4 Æ 61 1 Æ 77 12 Æ 00 0 Æ 002Stereotyped HCDR3 7 Æ 10 2 Æ 87 17 Æ 62 <0 Æ 001 4 Æ 45 1 Æ 67 11 Æ 86 0 Æ 003 IGHV4-39   usage 12 Æ 79 4 Æ 66 35 Æ 12 <0 Æ 0001 6 Æ 13 1 Æ 95 19 Æ 19 0 Æ 002HR, hazard ratio; LCI, lower confidence interval; UCI, upper confidence interval;  IGHV  , immunoglobulin heavy chain variable gene; HCDR3,immunoglobulin heavy chain complementarity determining region 3. Host Genetic Background and Risk of Richter Transformation ª  2010 Blackwell Publishing Ltd,  British Journal of Haematology  ,  152,  284–294  287  cumulative risk of RS transformation and to classify patientsinto risk categories. A model for RS prediction based onHCDR3 status and  IGHV4-39   usage (Rossi  et al  , 2009a) wasapplied to the RS cohort of this study. According to thismodel, RS were stratified into three risk categories (Fig 2A): (i)CLL utilizing both  IGHV4-39   and stereotyped HCDR3,characterized by a high risk of RS transformation (HR:43 Æ 96; 95% CI: 12 Æ 57–153 Æ 65; Events/  N  : 4/5; 5-year risk: 70 Æ 0%; P   < 0 Æ 001), (ii) CLL utilizing stereotyped HCDR3 but without IGHV4-39  , characterized by an intermediate risk of RStransformation (HR: 5 Æ 40; 95% CI: 2 Æ 05–14 Æ 19; Events/  N  : 10/72; 5-year risk: 13 Æ 7%;  P   = 0 Æ 001) and (iii) CLL withoutstereotyped HCDR3, independent of   IGHV4-39   usage, char-acterized by a low risk of RS transformation (Events/  N  : 7/429;5-year risk: 3 Æ 5%; reference category).The introduction of the  LRP4  rs2306029 genotype inthis model improved stratification by adding prognosticinformation within the intermediate risk group of patientscharacterized by stereotyped HCDR3 but without  IGHV4-39  .In particular, the  LRP4  rs2306029 genotype split intermedi-ate risk patients into two subgroups (Fig 2B): (i) CLLutilizing stereotyped HCDR3 without  IGHV4-39   andharbouring the  LRP4  rs2306029 TT genotype, characterizedby a high risk of RS transformation (HR: 14 Æ 13; 95% CI:4 Æ 72–42 Æ 29; Events/  N  : 6/16; 5-year risk: 30 Æ 4%;  P   < 0 Æ 001),(ii) CLL utilizing stereotyped HCDR3 without  IGHV4-39  and harbouring the  LRP4  rs2306029 CT/CC genotype,characterized by a low risk of RS transformation (HR:2 Æ 78; 95% CI: 0 Æ 81–9 Æ 53; Events/  N  : 4/56; 5-year risk: 8 Æ 5%%; P   = 0 Æ 103). Fig 1.  Cumulative probability of transformation to Richter syndrome.Kaplan–Meier plots representing the cumulative probability of trans-formation to Richter syndrome according to  LRP4  rs2306029 geno-type. Table III.  Biological and clinical characteristics of the CLL series according to  LRP4  rs2306029.Characteristics LRP4  rs2306029TT CT/CC P  N  /total % 95% CI (%)  N  /total % 95% CI (%)Biological variables IGHV   homology   ‡ 98 23/77 29 Æ 9 19 Æ 6–40 Æ 0 102/249 41 Æ 0 34 Æ 8–47 Æ 0 0 Æ 080Stereotyped HCDR3 19/75 25 Æ 3 15 Æ 4–35 Æ 1 58/249 23 Æ 3 18 Æ 0–28 Æ 5 0 Æ 716 IGHV4-39   usage 4/77 5 Æ 2 0 Æ 2–10 Æ 1 5/249 2 Æ 0 0 Æ 2–3 Æ 7 0 Æ 223FISH stratificationNormal 21/77 27 Æ 3 17 Æ 3–37 Æ 2 68/243 28 Æ 0 22 Æ 3–33 Æ 6 0 Æ 112del13q14 23/77 29 Æ 9 19 Æ 6–40 Æ 0 97/243 39 Æ 9 33 Æ 7–46 Æ 0+12 18/77 23 Æ 4 13 Æ 9–32 Æ 8 37/243 15 Æ 2 10 Æ 7–19 Æ 7del11q22-q23 3/77 3 Æ 9 0–8 Æ 2 19/243 7 Æ 8 4 Æ 4–11 Æ 2del17p13 12/77 15 Æ 6 7 Æ 4–23 Æ 6 22/243 9 Æ 1 5 Æ 4–12 Æ 6 TP53  mutation 7/76 9 Æ 2 2 Æ 7–15 Æ 7 25/230 10 Æ 9 6 Æ 8–14 Æ 8 0 Æ 682CD38  ‡ 30 20/78 25 Æ 6 15 Æ 9–35 Æ 3 84/248 33 Æ 9 27 Æ 9–39 Æ 7 0 Æ 174Clinical variablesAge >65 years 44/79 55 Æ 7 44 Æ 7–66 Æ 6 149/252 59 Æ 1 53 Æ 0–65 Æ 2 0 Æ 589Male 45/79 57 Æ 0 46 Æ 0–67 Æ 8 130/252 51 Æ 6 45 Æ 4–57 Æ 7 0 Æ 404Binet stage B–C 18/79 22 Æ 8 13 Æ 5–32 Æ 0 57/252 22 Æ 6 17 Æ 4–27 Æ 7 0 Æ 976Nodal areas  ‡ 3 11/79 13 Æ 9 6 Æ 2–21 Æ 5 41/252 16 Æ 3 11 Æ 7–20 Æ 8 0 Æ 617Lymph node size  ‡ 3 cm 9/79 11 Æ 4 4 Æ 3–18 Æ 4 15/252 9 Æ 9 3 Æ 0–8 Æ 8 0 Æ 707Absolute lymphocyte count >10  ·  10 9 /l 23/79 29 Æ 1 19 Æ 9–39 Æ 1 61/252 24 Æ 2 18 Æ 9–29 Æ 5 0 Æ 382Hb <100 g/l 3/79 3 Æ 8 0–8 Æ 2 14/252 5 Æ 6 2 Æ 7–8 Æ 3 0 Æ 771Platelets ( · 10 9 /l) 8/79 10 Æ 1 3 Æ 4–16 Æ 7 20/252 7 Æ 9 4 Æ 6–11 Æ 2 0 Æ 542LDH > ULN 12/79 15 Æ 2 7 Æ 2–23 Æ 1 25/252 9 Æ 9 6 Æ 2–13 Æ 6 0 Æ 195 IGHV  , immunoglobulin heavy chain variable gene; HCDR3, immunoglobulin heavy chain complementarity determining region 3; LDH, lactatedehydrogenase; ULN, upper limit of normal; FISH, fluorescence  in situ  hybridization. S. Rasi  et al  288  ª  2010 Blackwell Publishing Ltd,  British Journal of Haematology  ,  152,  284–294
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks