Letters to the Editor 1188 In this study, we describe a novel fusion in hematological malig- 3 Wasag B, Lierman E, Meeus P, Cools J, Vandenberghe P. The kinase inhibitor nancies, expanding number of FGFR1 partner and confirming TKI258 is active against the novel CUX1-FGFR1 fusion detected in a patient with the relevance of RANBP2/NUP358 in myeloid neoplasm. The clinical T-lymphoblastic leukemia/lymphoma and t(7;8)(q22;p11). Haematologica 2011; presentation as a myeloproliferative/myelodysplastic disorder was 96: 922–926. not different from other EMS. More cases with 2 4 Jackson CC, Medeiros LJ, Miranda RN. 8p11 myeloproliferative syndrome: a abnormalities should be explored in order to evaluate the incidence review. Hum Pathol 2010; 41: 461–476. 5 Shaffer LG, Slovak ML, Campbell LJ. An International System for Human Cytogenetic of RANBP2/NUP358 rearrangements in hematological neoplasms. Nomenclature. Karger: Basel, 2009. 6 Ma Z, Hill DA, Collins MH, Morris SW, Sumegi J, Zhou M et al. Fusion of ALK to the Ran-binding 2 (RANBP2) in inflammatory myofibroblastic tumor. CONFLICT OF INTEREST Genes Cancer 2003; 37: 98–105. The authors declare no conflict of interest. 7 Hidalgo-Curtis C, Chase A, Drachenberg M, Roberts MW, Finkelstein JZ, Mould S et al. The t(1;9)(p34;q34) and t(8;12)(p11;q15) fuse pre-mRNA processing SFPQ (PSF) and CPSF6 to ABL and FGFR1. Genes Chromosomes Cancer 2008; 47: ACKNOWLEDGEMENTS 379–385. 8 Soler G, Nusbaum S, Varet B, Macintyre EA, Vekemans M, Romana SP et al. We thank Dr Elisabeth FLORI for providing part of the BACs. LRRFIP1, a new FGFR1 partner gene associated with 8p11 myeloproliferative syndrome. Leukemia 2009; 23: 1359–1361. C Gervais1,2,3, L Dano1,2,3, N Perrusson4,CHe´lias1,2,3, E Jeandidier2,3, 9 Wu J, Matunis MJ, Kraemer D, Blobel G, Coutavas E. Nup358, a cytoplasmically A-C Galoisy1,3, A Ittel1,2,3, R Herbrecht5, K Bilger5 and exposed nucleoporin with peptide repeats, Ran-GTP binding sites, zinc fingers, a L Mauvieux1,2,3,4 cyclophilin A homologous domain, and a leucine-rich region. J Biol Chem 1995; 1Poˆle de Biologie, Laboratoire d’He´matologie, Hoˆpitaux Universitaires 270: 14209–14213. de Strasbourg, Strasbourg, France; 10 Bernad R, van der Velde H, Fornerod M, Pickersgill H. Nup358/RanBP2 attaches to 2 the nuclear pore complex via association with Nup88 and Nup214/CAN and plays Laboratoire Re´gional de Cytoge´ne´tique He´matologique d’Alsace a supporting role in CRM1-mediated nuclear protein export. Mol Biol 2004; 24: Mulhouse–Strasbourg, Strasbourg, France; 2373–2384. 3 Plateforme Re´gionale INCa de Ge´ne´tique Mole´culaire des Cancers 11 Patel AS, Murphy KM, Hawkins AL, Cohen JS, Long PP, Perlman EJ et al. RANBP2 d’Alsace, Strasbourg, France; and CLTC are involved in ALK rearrangements in inflammatory myofibroblastic 4Laboratoire d’He´matologie Cellulaire, Faculte´ de Me´decine de tumors. Cancer Genet Cytogenet 2007; 176: 107–114. Strasbourg, Strasbourg, France and 12 Chen ST, Lee JC. An inflammatory myofibroblastic tumor in liver with 5De´partement d’Onco-He´matologie, Hoˆpitaux Universitaires de ALK and RANBP2 gene rearrangement: combination of distinct morphologic, Strasbourg, Strasbourg, France, immunohistochemical, and genetic features. Hum Pathol 2008; 39: E-mail: [email protected] 1854–1858. 13 Rottgers S, Gombert M, Teigler-Schlegel A, Busch K, Gamerdinger U, Slany R et al. ALK fusion genes in children with atypical myeloproliferative leukemia. Leukemia 2010; 24: 1197–1200. REFERENCES 14 Aslanukov A, Bhowmick R, Guruju M, Oswald J, Raz D, Bush RA et al. RanBP2 1 Swerdlow SH, Campo E, Harris NL, Jaffe´ ES, Pileri SA, Stein H et al. WHO Classifi- modulates Cox11 and hexokinase I activities and haploinsufficiency of RanBP2 cation of Tumours of Haematopoietic and lymphoid tissues. IARC: Lyon, 2008. causes deficits in glucose metabolism. PLoS Genet 2006; 2:e177. 2 Li F, Zhai YP, Tang YM, Wang LP, Wan PJ. Identification of a novel partner gene, 15 Graux C, Cools J, Melotte C, Quentmeier H, Ferrando A, Levine R et al. Fusion of TPR, fused to FGFR1 in 8p11 myeloproliferative syndrome. Genes Chromosomes NUP214 to ABL1 on amplified episomes in T-cell acute lymphoblastic leukemia. Cancer 2012; 51: 890–897. Nat Genet 2004; 36: 1084–1089.

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Exome sequencing in tracking clonal evolution in multiple myeloma following therapy

Leukemia (2013) 27, 1188–1191; doi:10.1038/leu.2012.287 presents as a highly heterogeneous disease, this introduces a further complexity, requiring each subset of the disease to be evaluated separately in its response to therapy. Multiple myeloma (MM), defined as accumulation of malignant plasma cells in Sequencing the tumor genome using next-generation sequencing the bone marrow, exemplifies such a tumor and exhibits marked (NGS) is providing an unparalleled insight into the pathogenesis disease heterogeneity at presentation.1,2 At the outset, aberrant and progression of the disease. Much of the current focus of chromosomal markers exist in diagnostic MM samples that are NGS in cancer is on defining mutations in the tumor genome at associated with poor prognosis and help in delineating specific disease presentation, and these findings are central to under- disease subsets, including t(4;14) and 1q21 amplification among standing the molecular mechanisms that underlie pathogenesis. others,1–5 suggesting that these markers may be highly relevant to However, additional questions remain about the stability of the defining specific patterns of progression and clonal response to tumor clone under therapeutic pressure and these require a therapy. distinct second wave of NGS analyses. Understanding how each The pivotal NGS study of MM centered on the MM Research tumor evolves following therapy will be the key to delivering Consortium (MMRC) analysis of 38 tumor genomes that identified targeted therapy tailored for individual patients and in develop- key somatic mutations, either at disease presentation or at relapse, ing stratified therapeutic programs. Where a given tumor but not in a paired setting.6 Consequently, the data did not permit

Accepted article preview online 9 October 2012; advance online publication, 13 November 2012

Leukemia (2013) 1172 – 1218 & 2013 Macmillan Publishers Limited Letters to the Editor 1189 Table 1. Summary of sequence coverage metrics for three exomes from the analysis of a multiple myeloma index case displaying 1q21 amplification

Germline Presentation Y4

Read sequences 165002128 138654948 145003530 Aligned 156509645 130761431 137587296 % aligned 94.85310699 94.30707875 94.88548037 Mapped to target reads (%) 61.93 58.92 60.66 Mapped to target reads ±150 bp (%) 69.32 68.45 67.67 Mean coverage 136.63 109.14 120.48 Target bases with coverage 420 (%) 85.6 83.03 84.24

*Sequence capture: SureSelect Human All Exon 50M (Agilent), Sequencing: Illumina HiSeq. identification of molecular changes under therapy in individual capture next-generation sequencing with matched T-cell DNA for patients longitudinally. Subsequent NGS data described further germline. Sequencing libraries were subjected to Illumina/Solexa genomic lesions in MM that are associated with defined chromo- NGS (Supplementary Data). Bioinformatic comparison of tumor somal translocations, including t(4;14), and defined subclonal exomes with germline identified non-synonymous coding evolution,7 but again only at disease presentation. Importantly, no variants, representing putative mutations, with a summary of single gene mutation or combination of mutations has emerged sequence coverage metrics for the three exomes shown in as universal to all MMs at presentation, across tumor subsets, Table 1. A tumor sample was also available after conventional suggesting that multiple molecular pathways will most likely be induction chemotherapy in plateau phase (year 1; Y1), which was responsible for the pathogenesis of this disease and further only examined for selected gene mutations using conventional suggesting that clonal changes under therapy are likely to vary Sanger sequencing (Supplementary Data; Table 2b). At Y4, routine between disease subsets in MM, to be mapped fully. chromosomal assessment had already revealed the emergence of Data are beginning to emerge to meet this mapping need, but t(11;14) in 30% of tumor cells as a progression event at markedly different levels of molecular sensitivities, determined (Supplementary Data). by the technologies used. Very recently, a study based on array By coding exome capture next-generation sequencing results comparative genomic hybridization analysis on the 244A Human with 0 or 1 to denote the absence or presence of apparent variant Genome CGH Microarrays (Agilent) platform evaluated 28 MM (Table 2a), 81 tumor-specific novel non-synonymous variants were cases treated with a range of therapeutic regimens, in paired identified, of which 33 were shared at presentation and Y4 and longitudinal analyses in individual patients.8 This specific 48 were de novo variants at Y4 (and Supplementary Table S1). technology is inherently limited, defining insertions and Of the 81 genes, 14 were mutated in the MMRC cohort, with NRAS deletions of large DNA fragments of 4several kb size and (p.Q61K) being identical,6 and all but one gene (WDR73) were also assessing copy number variation or copy number abnormality to mutated in COSMIC (two identical: NRAS,p.Q61K;TAF1L,p.R1243W) define allele gain or loss. This study by Keats et al. reported (Supplementary Table S1). To verify fidelity, B20% of novel varying patterns of clonal response to therapy in MM, with some variants from each 0,1,1 and 0,0,1 data set were randomly Sanger cases displaying a relatively stable clone and others showing sequenced to reveal a correct positive call of 100% (Table 2a and differing patterns of marked genetic changes.8 Nevertheless, the b), suggesting high confidence in exome capture next-generation array comparative genomics hybridization-based study sequencing bioinformatic calls. Of 15 confirmed mutations, B50% established a necessary ‘broad-brush’ understanding of clonal are predicted to be damaging (Table 2b). Notably, these findings response to therapy in MM, observing that patients with poor- reveal and confirm clonal evolution in MM following therapy, in prognosis chromosomal markers tend to be associated with the our case accruing mutations at presentation that have evolved largest degree of genetic change.8 during the course of disease and persist to Y4, and acquiring Defining the precise pathways of response to therapy in MM mutations by Y4 that have arisen after plateau phase (Y1; cells, however, will require a concise mapping of somatic mutations Table 2b). They parallel the very recent observations from serially at the nucleotide level, at the next layer of molecular sensitivity, tracking the single t(4;14) MM case from presentation to end-stage in coding and non-coding genome-wide regions, and this will be plasma cell leukemia using whole-genome sequencing.9 Therapy dependent on NGS. Mutated genes determine altered effector differed in that study, including lenalidomide initially followed by molecule functions and mediate mechanisms of relapse and a bortezomib combination. In the former study, 10 variants were resistance to therapy. These mutations will need to be defined in common throughout and other clonal variants appeared and specific subsets of MM after therapy. This level of NGS analysis has disappeared during disease course, with five gene variants very recently been used to examine a single MM case displaying a associating with transformation to plasma cell leukemia.9 Two t(4;14) abnormality, at different time points pre-and post-therapy.9 mutated genes are common to both studies, CSMD3 (CUB and These data revealed a clear clonal evolution as disease progresses Sushi multiple domains 3) in presentation and relapse, and SUB1 following therapy and defined key mutations in specific genes. (SUB1 homolog, S. cerevisiae) associating with Y4 and plasma cell We extend the use of NGS in tracking the tumor clone in MM leukemia stage,9 suggesting important roles in tumor progression. after therapy, by examining a separate subset of disease. By However, another gene AFF1 (AF4/FMR2 family, member 1) was focusing on NGS of the tumor exome, we examine somatic mutated at presentation to the plasma cell leukemia phase in the mutations that have a direct impact on gene-coding sequences former study,9 and was highlighted as possibly relevant to and report on the exome in an index case of MM with 1q21 progression in t(4;14) MM but was absent in our study, pointing amplification, comparing disease presentation with relapse in a to possible disease subset variation. Potential divergence at Y4 paired analysis. (48 variants) in the present study is large, however, as only 33 The MM case presented as IgGk stage IIIA and underwent variants appear at presentation following pathogenesis, suggest- phases of treatment including chemotherapy and autologous ing stage-dependent kinetic differences in acquiring mutations transplants (described further in Supplementary Data). Tumor that associate with tumor progression in response to therapy. DNA was examined at disease presentation and at first relapse The present report and the study by Egan et al.9 delineate the after the start of first-line therapy (year 4; Y4), using exome nature of individual gene mutations that occur following therapy

& 2013 Macmillan Publishers Limited Leukemia (2013) 1172 – 1218 1190 ekma(03 12–1218 – 1172 (2013) Leukemia

Table 2. Novel non-synonymous variants identified in tumor exomes at presentation and relapse in an index multiple myeloma case with 1q21 amplification

(a) Summary

Category Total Known Novel Validated novel variants

0,1,0 0 0 0 n/a 0,1,1 39 6 33 6/6 0,0,1 48 0 48 9/9

(b) Validated variants by Sanger DNA sequencing

Category Chr hg18 Gene Nucleotide Amino acid MMRF COSMIC Egan Germline Presentation Y4 Presence/absence Functional basepair cohort6 et al.9 alternative alternative alternative of variant prediction

allele reads allele reads allele reads confirmed Editor the to Letters (%) (%) (%) in four samplesx

0,1,1 4 187755248 FAT1y C4A p.G4147V N Y N 0 (0) 27 (34.18) 40 (53.33) 0,1,1,1 Probably damaging 0,1,1 10 11247564 CELF2 G4A p.E62K, p.E55K, N Y N 0 (0) 31 (46.27) 38 (46.91) 0,1,1,1 Benign p.E31K 0,1,1 10 71994192 KIAA1274 C4T p.R777C N Y N 0 (0) 15 (41.67) 21 (33.87) 0,1,1,1 Probably damaging 0,1,1 8 113401311 CSMD3 T4C p.E2414G, Y Y Y 0 (0) 76 (47.5) 67 (43.23) 0,1,1,1 Probably p.E2374G, damaging p.E2310G 0,1,1 7 91470201 AKAP9 T4C p.C1012R N Y N 0 (0) 74 (32.89) 84 (33.6) 0,1,1,1 Benign 0,1,1 5 140735725 PCDHGA6 G4A p.A631T N Y N 0 (0) 36 (36) 27 (29.67) 0,1,1,1 Benign/ possibly damaging 0,0,1 6 35812968 C6orf81 C4G p.I35M N Y N 0 (0) 0 (0) 15 (23.81) 0,0,0,1 Probably damaging 0,0,1 5 32822145 NPR3 T4G p.S304R, p.S520R, N Y N 0 (0) 0 (0) 67 (22.95) 0,0,0,1 Benign p.S521R 0,0,1 15 82989852 WDR73 G4A p.S246L N N N 0 (0) 0 (0) 20 (23.26) 0,0,0,1 Possibly damaging 0,0,1 11 107011595 ELMOD1 G4C p.C105S N Y N 0 (0) 1 (0.19) 148 (24.79) 0,0,0,1 Benign 0,0,1 7 123043524 ASB15 A4G p.S41G Y Y N 0 (0) 0 (0) 43 (23.89) 0,0,0,1 Benign & 0,0,1 2 227857236 COL4A3 G4A p.D938N Y Y N 0 (0) 0 (0) 19 (29.69) 0,0,0,1 Benign 03McilnPbihr Limited Publishers Macmillan 2013 0,0,1 11 22734034 GAS2 C4A p.Q228K N Y N 0 (0) 0 (0) 20 (20.2) 0,0,0,1 Benign 0,0,1 5 32634839 SUB1 T4C p.Y71H N Y Y 0 (0) 0 (0) 35 (23.81) 0,0,0,1 Possibly damaging 0,0,1 1 48633603 SPATA6 G4C p.P264R N Y N 0 (0) 0 (0) 26 (44.83) 0,0,0,1 Probably damaging

Abbreviation: COSMIC, Catalog of Somatic Mutations in Cancer; MMRC, Multiple Myeloma Research Consortium; Novel ¼ not present in dbSNP129, dbSNP132, The 1000 Genomes Project and the NHLBI GO Exome Sequencing Project (ESP). All novel variants are described in Supplementary Table 1, with full bioinformatic metrics shown in Table 1. Category (Panel a) describes for absence (using notation 0) or presence (1) of variant by bioinformatic call in exome order: germline, presentation sample, Y4 sample. Number of randomly selected variants validated by Sanger DNA sequencing in each Category (Panel a): 6 in 0,1,1 and 9 in 0,0,1. Functional predictions: PolyPhen-2 v2.2.2r398. xPanel b: Absence (0) or presence (1) of variant as confirmed by Sanger DNA sequencing in 4 samples ordered as: germline, presentation sample, Y1 plateau phase sample, Y4 relapse sample. yPanel b: listed as a possible false-positive gene. Letters to the Editor 1191 in MM in two distinct poor prognosis subsets and have initiated 4First Department of Medicine, Center for Oncology and Hematology, the phase of NGS work that will be necessary to elucidate at the Wilhelminenspital, Vienna, Austria and gene level the ‘broad-brush’ insight delivered by the study by 5The GenePool, Ashworth Laboratories, University of Edinburgh, Keats et al.8 Taken together, our observations and those from the Edinburgh, UK single t(4;14) case9 reveal a disease landscape at relapse in MM E-mail: [email protected] that displays complex patterns of genetic mutations that will need 6Joint first authors in this study. 7 to be dissected by NGS in relation to specific therapies. Common Joint senior authors in this study. mutations, such as CSMD3 and SUB1 as described above, may provide key pointers to essential survival pathways irrespective of type of therapy or disease subset. Whether specific patterns associate with individual markers of poor prognosis in MM will REFERENCES emerge from NGS studies of larger cohorts. These patterns will 1 Carrasco DR, Tonon G, Huang Y, Zhang Y, Sinha R, Feng B et al. High-resolution define the models of clonal evolution, which may be linear or may genomic profiles define distinct clinico-pathogenetic subgroups of multiple involve a clone regressing and reappearing by competition. myeloma patients. Cancer Cell 2006; 9: 313–325. 2 Chesi M, Bergsagel PL. Many multiple myelomas: making more of the molecular Whether all 1q21 MM cases exhibit an apparent linear mode of mayhem. Hematology Am Soc Hematol Educ Program 2011; 2011: 344–353. evolution under therapy, as suggested by our data, remains to be 3 Hanamura I, Stewart JP, Huang Y, Zhan F, Santra M, Sawyer JR et al. Frequent gain determined. of chromosome band 1q21 in plasma-cell dyscrasias detected by fluorescence It is however evident from these two initial NGS studies that the in situ hybridization: incidence increases from MGUS to relapsed myeloma and is tumor clone will need to be closely scrutinized after each phase of related to prognosis and disease progression following tandem stem-cell trans- treatment. These studies also suggest that a multi-target plantation. Blood 2006; 108: 1724–1732. combination therapy may be required to eradicate the genetically 4 Klein U, Jauch A, Hielscher T, Hillengrass J, Raab MS, Seckinger A et al. variant subclones in any specific tumor population in MM that Chromosomal aberrations þ 1q21 and del(17p13) predict survival in patients with persist after early therapy to prevent escape.10 recurrent multiple myeloma treated with lenalidomide and dexamethasone. Cancer 2011; 117: 2136–2144. 5 Kumar SK, Mikhael JR, Buadi FK, Dingli D, Fonseca R, Gertz MA et al. Management CONFLICT OF INTEREST of newly diagnosed symptomatic multiple myeloma: updated mayo Stratification of Myeloma and Risk-Adapted Therapy (mSMART) consensus guidelines. Mayo The authors declare no conflict of interest. Clinic Proc 2009; 84: 1095–1110. 6 Chapman MA, Lawrence MS, Keats JJ, Cibulskis K, Sougnez C, Schinzel AC et al. ACKNOWLEDGEMENTS Initial genome sequencing and analysis of multiple myeloma. Nature 2011; 471: 467–472. This work was funded by EU FP7 Program Project 278706 ‘OVER-MyR’, Leukemia & 7 Walker BA, Wardell CP, Melchor L, Hulkki S, Potter NE, Johnson DC et al. Intraclonal Lymphoma Research (UK) and Cancer Research UK (JG). heterogeneity and distinct molecular mechanisms characterize the development of t(4;14) and t(11;14) myeloma. Blood 2012; 120: 1077–1086. 8 Keats JJ, Chesi M, Egan JB, Garbitt VM, Palmer SE, Braggio E et al. Clonal compe- 1,6 2,6 3 2 4 N Weston-Bell , J Gibson , M John , S Ennis , S Pfeifer , tition with alternating dominance in multiple myeloma. Blood 2012; 120: T Cezard5, H Ludwig4, A Collins2,7, N Zojer4,7 and SS Sahota1,7 1067–1076. 1Tumour Immunogenetics Group, Cancer Sciences Academic Unit, 9 Egan JB, Shi CX, Tembe W, Christoforides A, Kurdoglu A, Sinari S et al. Whole Faculty of Medicine, University of Southampton, Southampton, UK; genome sequencing of multiple myeloma from diagnosis to plasma cell leukemia 2Genetic Epidemiology and Genomic Informatics Group, reveals genomic initiating events, evolution and clonal tides. Blood 2012; 120: Human Development and Health Academic Unit, 1060–1066. 10 Al-Lazikani B, Banerji U, Workman P. Combinatorial drug therapy for cancer in the Faculty of Medicine, University of Southampton, post-genomic era. Nat Biotechnol 2012; 30: 679–692. Southampton, UK; 11 Fuentes Fajardo KV, Adams D, Mason CE, Sincan M, Tifft C. NISC Comparative 3 Department of Preclinical Sciences, Faculty of Medical Sciences, Sequencing Program. Detecting false-positive signals in exome sequencing. Hum University of The West Indies, Trinidad and Tobago; Mutat 2012; 33: 609–613.

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Matriptase is highly upregulated in chronic lymphocytic leukemia and promotes cancer cell invasion

Leukemia (2013) 27, 1191–1194; doi:10.1038/leu.2012.289 Studies indicate that matriptase may have an important role in cancer. Matriptase overexpression has been reported in many human solid tumors, including prostate, breast, Matriptase, also called MT-SP1 or epithin, is a type-II transmem- ovarian, kidney and lung cancers.4,5 In mice, transgenic brane serine protease expressed on the surface of the normal matriptase expression in keratinocytes causes skin carcinoma.6 epithelium.1,2 The enzyme is required for epidermal differentiation The molecular mechanism under which matriptase may and barrier function in the skin. Low levels of matriptase mRNA participate in cancer is not fully understood. Matriptase has also were detected in human blood cells including monocytes and been shown to activate pro-urokinase, prostasin, protease-acti- B lymphocytes,3 but the biological significance of such expression vated receptor-2, pro-hepatocyte growth factor, pro-macrophage is unclear. In monocytes, matriptase may be involved in initiating stimulating protein-1, platelet-derived growth factor, vascular plasminogen activation.3 endothelial growth factor receptor 2 and Tie2 receptor.2,4,7–9 It is

Accepted article preview online 9 October 2012; advance online publication, 26 October 2012

& 2013 Macmillan Publishers Limited Leukemia (2013) 1172 – 1218