Leukemia (2001) 15, 1627–1632  2001 Nature Publishing Group All rights reserved 0887-6924/01 $15.00 www.nature.com/leu Chromosomal changes pattern and amplification in T cell non-Hodgkin’s lymphomas M Renedo3, B Martinez-Delgado1, E Arranz1,3, MJ Garcia2, M Urioste1, A Martinez-Ramirez1, C Rivas2, JC Cigudosa1 and J Benitez1

1Department of Human Genetics, Centro Nacional de Investigaciones Oncologicas, Madrid, Spain; 2Department of Pathology, Fundacio´n Jimenez Dı´az, Madrid, Spain; and 3GEMOLAB, Madrid, Spain

T cell non-Hodgkin’s lymphomas are a heterogeneous group genetic techniques is that CGH does not need cell culture to of lymphomas with poor prognosis, and whose genetic alter- obtain metaphases and it allows a global genetic analysis of ations are not well understood. Comparative genomic hybridiz- ation (CGH) is a technique that allows the identification of DNA the genome by passing the examination of a minor cell clone imbalances without cytogenetic studies. We have studied 37 that was not representative of the tumoral genetic profile. samples from 29 T cell non-Hodgkin’s lymphomas (25 periph- We have performed CGH in 37 samples from 29 patients eral and four lymphoblastic lymphomas) by CGH in order to with Tcell NHL, in order to detect DNA sequence copy num- detect DNA sequence copy number changes of putative impor- ber changes along all . Our results point to spe- tance in the biology and prognosis of these neoplasms. We cific regions previously described by cytogenetic techniques, detected abnormal CGH profiles in 16/27 (59%) of samples at diagnosis, a ratio that increased to 66% (23/37) when we as well as other new regions not previously described. Ampli- included the relapsed samples. The most common recurrent fication of some mapped in one of these regions, changes were gains related to the X , either the detected by Southern and FISH analysis was concordant with whole chromosome or partially the Xq26–27 bands (19%). Other the CGH results. recurrent changes included gains of bands 9q34, gains of chro- mosomes 17, 19, and 20, and complete or partial deletions of chromosome 13 (10%). -related genes located at Xq26– 28 region were analyzed by Southern blot and fluorescence in Materials and methods situ hybridization (FISH). Low level amplification of some of these genes was detected by this technique confirming the Patients results obtained by CGH in this region. The detection of abnor- mal CGH profiles in these T cell lymphomas could have clinical High molecular weight DNA was extracted from 37 samples implications. Patients with abnormal CGH profiles showed sig- belonging to 29 patients diagnosed of T-NHL; 35 fresh lymph nificant associations with advanced stage of disease, over- expression of P53, and higher proliferative index. Leukemia nodes and two bone marrow samples. According to the WHO 7 (2001) 15, 1627–1632. classification the 29 Tcell NHL included 25 peripheral Tcell Keywords: T cell lymphomas; CGH; gains and losses; gene ampli- lymphomas (PTCL) (two angioimmunoblastic, three anaplastic fication large cell, two mycosis fungoides, 17 unspecified and one unclassifiable) and four precursor Tcell lymphoblastic lym- phomas. The 17 unspecified PTCL corresponded to the pleo- Introduction morphic and lymphoepithelioid types of the Kiel classi- fication.8 In seven cases, different samples were obtained at Tcell lymphomas are a heterogeneous group of non- different stages during the evolution of the disease. A summary Hodgkin’s lymphomas (NHLs) with a highly variable mor- of clinical data is shown in Table 1. phology and controversial classification. In Spain, these tumors are less common than those of B cell origin, as occurs in the rest of Europe (10–15% of all NHLs). Moreover, com- Immunohistochemistry pared with B cell neoplasms, Tcell NHLs appear to have a worse prognosis.1 Tissue was routinely fixed in 10% formalin, embedded in par- The genetic data on T cell neoplasms are still limited. affin and stained with hematoxylin-eosin (E–H) and Giemsa Although non-random abnormalities at the cytogenetic level in thin sections (4 ␮m). have been described, such as rearrangements involving An immunohistochemical study was performed with a 14q11, 14q32, 6p23, 1p36,2 and trisomies of chromosomes streptavidin-biotin peroxidase complex technique using a X, 3 and 7q,3–6 the molecular changes that could play a role in complete panel of monoclonal antibodies against Tand B tumoural pathogenesis and progression and in drug resistance lymphocytes (CD45, CD3, CD4, CD8, CD20), histiocytes remain unknown. (CD68), dendritic cells (CD21), activation markers (CD30, Comparative genomic hybridization (CGH) is a technique CD15, EMA) and aggressiveness (MIB1, P53) after heat pre- allowing the identification of imbalances of the entire genome treatment for antigenic retrieval. The two monoclonal anti- in terms of DNA copy number with a putative diagnostic and bodies used to consider neoplastic aggressivity, MIB1 and p53 prognostic value that could remain undetected by conven- showed nuclear strong positivity, without stained cytoplasms. tional cytogenetics. Its main advantage with respect to cyto- Immunocytochemistry technique with previous heat antigen retrieval avoids non-specific cytoplasmic discolored patterns. The quantification of MIBI and the expression of P53 was determined with an image analyzer (CAS-200, Becton Correspondence: J Benı´tez, Dept Gene´tica Humana, Centro Nacional Dickinson) with proliferation index program which provides de Investigaciones Oncolo´gicas Carlos III, Ctra Majadahonda-Pozuelo Km 2,200, 28220 Majadahonda, Madrid, Spain; Fax: 34 91 5097055 an accurate and objective method for assessment. P53 over- The first two authors contributed equally to this work expression was defined when positive cells were over 10%. Received 27 February 2001; accepted 15 June 2001 P53 staining is absent in reactive lymphoid tissue, and 10% Chromosomal and gene amplifications in T-NHL M Renedo et al 1628 Table 1 Clinical and tumor status data from 29 patients with Tcell NHL

Case Stage Status Sex Subtype Immunophenotype P53 PI % TCR CGH

1a IA R M PTCL (Pleomorphic) CD3+, CD4+++, CD8+ + 5MN 1b IVA 2nd R ND ND M A 2 NA D F PTCL (Pleomorphic) CD3+, CD4+++, CD8+ + 5PA 3 NA D F PTCL (Pleomorphic) CD3+, CD4+++, CD8+ − 15–20 P N 4 IA D F PTCL (Pleomorphic) CD3+, CD4+++, CD8+ + 10–30 P A 5 NA D F PTCL (Pleomorphic) CD3+, CD4+, CD8+ − 30 ND N 6 IVB D M PTCL (Pleomorphic) CD3+, CD4+++, CD8+ − 50 P A 7a IIIA D CD3+, CD4+++, CD8+ + 60 M A 7b IVA R M PTCL (Pleomorphic) ND 60 P A 7c IVB R ND 55 M A 8a NA D F PTCL (Pleomorphic) CD3+, CD4+++, CD8+, CD30+ − 20–30 M N 8b R + ND M A 9 IVB D F PTCL (Pleomorphic) CD3+, CD4++, CD8+, CD30+ − 50 M A 10 IVA D F PTCL (Pleomorphic) CD3+, CD4+++, CD8+, CD30+ ND 70 P A 11 NA D M PTCL (Pleomorphic) CD3+, CD4++, CD8+ ND 20 P N 12 NA D M PTCL (Pleomorphic) CD3+, CD4+++, CD8+, CD30+ − ND M N 13 NA D M PTCL (Lymphoepithelioid) CD3+, CD4+++, CD8+ + 5–10 M A 14 NA D F PTCL (Lymphoepithelioid) CD3+, CD4++, CD8+ − 20–30 M A 15a NA D M PTCL (Lymphoepithelioid) CD3+, CD4++, CD8+ − 15–20 ND N 15b R ND ND M N 16a IA D M PTCL (Lymphoepithelioid) CD3+, CD4++, CD8+, CD30+ − 30 M N 16b CR ND ND P N 17 NA D M PTCL (Lymphoepithelioid) CD3+, CD4+++, CD8+ ND 15–20 M N 18 NA D M PTCL (AILT) CD3+, CD4++, CD8+ − 80 M A 19 IIIB R M PTCL (AILT) CD3+, CD4++, CD8+ − 60 M A 20 IA D M PTCL (Anaplastic) CD3+, CD30+ + 30 M A 21 NA D M PTCL (Anaplastic) CD3+, CD30+, ALK1+ + ND P A 22a IIA D F PTCL (Anaplastic small CD3+, CD30+ − 80 M N 22b R cell variant) + ND M A 23 IIB D F PTCL (MF) CD3+, CD4++, CD8+ − 5–10 P A 24 NA D M PTCL (MF) CD3+, CD4++, CD8+ − 5PA 25 NA D M PTCL (Unclassifiable) ND ND M A 26a IVA D M Lymphoblastic CD3+, CD1+ − 80 M N 26b IVB R ND ND M N 27 IVA D F Lymphoblastic CD3+, CD1− + 70 P A 28 IVA D F Lymphoblastic CD3+, CD1+ − 80 ND A 29 NA D M Lymphoblastic CD3+, CD1+ ND 80 ND A

PTCL, peripheral T cell lymphoma; AILT, angioimmunoblastic T cell lymphoma; MF, mycosis fungoides; D, diagnosis; R, relapse; CR, complete remission; PI, proliferative index; +, P53 protein levels Ͼ10%; −, P53 protein levels Ͻ10%; ND, no data; TCR, T cell receptor rearrangements; M, monoclonal; P, polyclonal; NA, not available; A, altered CGH profile; N, normal CGH profile.

was accepted as a correct value of expression in high-grade Mannheim). The amplification cycles consisted on: 45 s at lymphomas. In relation to MIB1, we considered cases under 94°C, 45 s at 54 or 55°C (mix I or mix II) and 45 s at 72°C, and over 40% of positive cells to correlate with the presence followed by final extension of 5 min at 72°C. The PCR pro- of altered CGH profiles. There is not consensus about precise ducts were then resolved in a 9% polyacrylamide gel (19:1 quantitative values of MIB1 to differentiate low and high acrylamide:bisacrylamide ratio), and then stained with grades in Tcell lymphomas, but in our experience at over 40– ethidium bromide. 50% a major aggressiveness is expected.9 Estimates of the proportion of tumoral cells ranged from 60 to 100% in peripheral T-lymphomas and between 90 and 100% in lymphoblastic lymphomas. Comparative genomic hybridization

TCR (T cell receptor) gene rearrangements Comparative genomic hybridization (CGH) was performed according to Kallionemi et al,11 with some modifications. The TCR-gamma chain gene rearrangements were performed in all tumor (test) and normal (reference) DNAs were labelled with tumors. We used PCR with primers previously described10 to SpectrumGreen and SpectrumOrange fluorochromes by using detect clonal tumor cells in these samples. We carried out the nick translation kit from Vysis, IL, USA. All samples were two multiplex PCRs with the following primers mixes: mix I co-hybridized with sex-matched reference DNA. For each contained TCR V2, V3, V4, V8 and V9 and mix II contained group of experiments, positive and negative controls (cases TCR V5, V10, V11, and V12. For the J segments, three primers with and without known changes) were analyzed at the same were mixed JGT1–2, JGT3 and JGT4. PCR reactions were per- time. Briefly, 200 ng of each labelled DNA were hybridized formed in 50 ␮lof1× PCR buffer (50 mM KCl, 10 mM Tris to normal male metaphase spreads in the presence of 15 ␮g ␮ (pH 8.3), 1.5 mM MgCl2), 200 M of dNTPs, 25 pmol of each of Cot-1 DNA, for 3 days. After washes, chromosomes were primer, 250 ng of DNA, 2.5 U Taq polymerase (Boehringer counterstained with DAPI in an antifade solution.

Leukemia Chromosomal and gene amplifications in T-NHL M Renedo et al 1629 Digital image analysis and interpretation of CGH restriction enzymes, HindIII and EcoRI. After electrophoresis, results the DNA was transferred to a nylon membrane and then hybridized using radioactive labelled probes from genes All CGH experiments were performed twice and indepen- located at Xq26–28 (SH2D1A, MCF2, MAGE1 and MTCP1) dently by two of us. Slides were analyzed with a Nikon epi- obtained from the following IMAGE clones: 1032031, fluorescence microscope equipped with a charged coupled 1755974, 1757557 and 701515, respectively. First, these device camera. A minimum of seven to 10 metaphases per clones were grown and DNA was extracted following stan- hybridization and case were analyzed using two image analy- dard methods. The inserts were sequenced to confirm the cor- sis systems (Quips, Vysis; and Cytovision, Applied Imaging, responding sequence of each gene before they were used as UK) that can process three-color images, green for the tumor probes. After digestion with HindIII enzyme, hybridization DNA hybridization, red for the normal reference DNA with MCF2 probe generated two fragments of 4.6 and 1.8 kb, hybridization and blue for the DNA counterstain. and with SH2D1A probe, a 3.4 kb fragment. Digestion with The green and red fluorescence intensities were calculated EcoRI and hybridization with MAGE and MTCP1 probes gen- and the green-to-red ratio profiles along the chromosome axis erated fragments of 4.4 and 4.6 kb, respectively. The intensity were displayed. Ratios greater than 1.20 and less than 0.80 of bands was quantified by means of densitometric analysis. were considered to represent chromosomal gain and loss, To estimate the gained copy number, the results were standar- respectively. High level amplification was defined as occur- ized with the results obtained from a control gene (␤-actin) rence of fluorescence intensity values in excess of 2.0 and and compared to normal samples. The sex of the patients distinct strong localized (band-like) green signal. Telomeric included in the analysis was always considered in order to and heterochromatic regions were excluded from the determine the level of amplification. analysis.12

Statistical analysis DNA probes fluorescence in situ hybridization (FISH) We used the two coils Fisher exact test with a confidence Direct-fluoresceion isothiocyanate (FITC) and digoxine- interval of 95% for detecting any possible significant associ- labelled probes for the X centromere were cohybridized in ations between the clinical parameters and the CGH results. separate assays together with some of the same probes of the Xq26 regions that were used for the Southern blot: MAGE1, SH2D1A, MTCP1 and MCF2. These gene probes were first Results hybridized on to normal metaphase spreads to detect any chimeric signal and to confirm the localization of the signal All tumoral samples showed an immunophenotype compat- at Xq26 (data not shown). The probe for gene MAGE1 showed ible with Tcell lymphomas, and most of them (63%) also a clear and distinct signal and it was our choice for the tests. showed monoclonal rearrangements of the TCR-gamma chain Diagnostic FISH assays were then done on paraffin-embedded gene (Table 1). tissue. Briefly, 5 ␮m sections were cut from the paraffin block CGH analysis revealed a 59% of samples at diagnosis and placed on to silanized slides. The tissues were deparaffin- (16/27 cases) with chromosomal aberrations. This percentage ized using two changes of xylene and two changes of 100% increased to 66% when the additional relapse samples were ethanol, both 10 min each. The samples were then pretreated included (23/37 samples) (Table 1). The most frequent with a solution of sodium bisulfite (1%) in 2 × SSC, and changes included gains at chromosome X (19%), 19 (10%), digested with 10 mg/ml of proteinase K (at 40°C and during 17 (10%), 20 (10%) and gain of 9q34 band (10%) and com- 40 min each wash). Denaturation was done in 70% plete loss or partial deletion of chromosome 13 (10%). Other formamide/2 × SSC for 2 min at 72°C. The denatured slides less common alterations were specific gains of 11q13 (8%), were dehydrated in cold ethanol. The probe mixture was and 5p15 band (8%) and gains at chromosome 7 with the denatured at 75°C for 10 min, applied to the slides, covered commonly gained region at 7q32-ter (8%) (Figure 1). and sealed with glass coverslips, and incubated overnight at 37°C. Two washing steps were carried out with 0.4 × SSC/0.1% Nonidet-P40 at 45°C, and, when necessary, detec- tion was performed with avidin FITC anti-digoxigenin sol- ution. The samples were counterstained with DAPI. Hybridization signals were visualized using an Olympus microscope, and images were captured on a CCD camera, using Cytovision image software. Two hundred nuclei that showed both green (X centromere) and red (genes mapped at Xq26) signals were counted.

Southern blot hybridization

In order to confirm these gains at the molecular level, South- ern blot was performed in three tumors that had gains of the entire or specific gains of the Xq26–27 region Figure 1 Summary of the main chromosome gains (vertical lines to the right of the chromosomal ideograms) and losses (vertical lines as revealed by CGH. Two tumor samples with normal CGH to the left of the chromosomal ideograms) detected by CGH. Thick profiles, and two normal peripheral blood samples were also bars indicate high level amplification. The number at the top of each analyzed. DNA was digested separately with two different bar represents the number of the tumour.

Leukemia Chromosomal and gene amplifications in T-NHL M Renedo et al 1630 Ten of 19 patients showed multiple aberrations in different chromosomes, but a single alteration was found in the remain- ing nine patients. These single alterations included the gain of the X chromosome (case 1), gain of the 3q27-qter region (case 2), gains at 7q (case 27) and two lymphomas (cases 6 and 8) that shared the gain of 5p15 region as the only alteration at diagnosis. Other single alterations were found in some lym- phomas, but they constituted unique events not shared by other tumors. In seven cases, serial samples were obtained periodically. Changes in the CGH profile were observed in four of them (cases 1, 7, 8, 22). Patients 1 and 22 showed first a normal CGH profile, but the gain of whole chromosome X appeared in the second relapse of patient 1 and multiple alterations were found in the relapse sample of patient 22. In patient 7, the gain of the X chromosome was present at diagnosis and was maintained in two relapse samples. In case 8, gain of the 5p15 band was detected at diagnosis, and additional changes were observed at relapse. In the other patients (cases 15, 16, 26) a normal profile was found in the different stages of the disease. In two lymphomas (cases 22 and 25) we could detect high level amplifications. Case 22 showed a very complex CGH profile at relapse that included gains and losses at different chromosomes and clear amplification signal at 17q12–q21 bands (Figure 2). Case 25 presented amplification of 19q13.1 band. Figure 2 (a) CGH profile of the X chromosome of case 7a showing the gain of the whole chromosome, and a signal of amplification on the Xq26–28 region. (b) CGH profile of the chromosome 17 of case Association of CGH results with the histological and 22 showing a gain of the q arm, and a signal of high level amplifi- clinical features cation on bands 17q12–21. (c) FISH image of a paraffin section of case 7a. The sample corresponded to a male patient. 35% of the The presence of chromosomal alterations detected by CGH in nuclei showed single green and red signals: one X chromosome and our series does not seem to be associated with any specific one copy of the MAGE1 gene, respectively. The remaining nuclei subtype of T cell lymphomas (Table 1). Two thirds of both showed two green signals (gain of X chromosome) and three to five red signals, corresponding to the gains of the MAGE1 gene. An abnor- types of peripheral Tcell and lymphoblastic lymphomas harb- mal cell is shown in the right side of the picture: white arrows point ored genetic changes. We did not find CGH gains or losses to the amplification signal of the MAGE1 probe in a cell that also in the two cases of mycosis fungoides analyzed. In addition, shows gain of chromosome X (two signals). Other cells with single the most relevant aberrations were found in almost all sub- copies of chromosome X and the MAGE1 probe are also displayed. types analyzed. We tried to correlate CGH results statistically with the fol- lowing clinical data: clinical stage, histological subtype, over- Table 2 Clinical and prognostic parameters Tcell NHLs with expression of p53, and PI. An abnormal CGH profile was asso- normal and abnormal CGH profiles ciated with more advance clinical stages (P = 0.03389), with overexpression of P53 (P = 0.037), and with PI over 40% (P Clinical and prognostic Normal Abnormal P values = 0.03661) (Table 2). No statistically significant association parameters CGH CGH was found for the histological subtype. profile profile

Clinical stage n = 6 n = 12 0.03389 Stages I and II 4 2 Gene and FISH analysis Stages III and IV 2 10 Southern blot analysis of some genes located in Xq26–28, P53 overexpression n = 9 n = 17 0.037 which is the minimal region involved in gains affecting chro- 19 mosome X in these lymphomas, confirmed the presence of Proliferative index n = 10 n = 18 0.003661 gains of genetic material at the molecular level. We selected Ͻ40% 8 7 Ͼ MCF2, SH2D1A, MAGE1 and MTCP1 genes because of their 40% 2 11 possible implication in different neoplasias. Three of the cases (tumour 7 at diagnosis, tumour 7 at relapse, and tumour 25) n, number of patients in whom clinical and prognostic parameters could be obtained. with gain of this region as revealed by CGH, showed gains of three of the analyzed genes, SH2D1A, MAGE1 and MTCP1 by densitometric analysis (Figure 3). However, no amplifi- cations of any of these genes were found in tumor samples patient) using the X-centromere probe and the MAGE1 probe with normal CGH profiles or normal peripheral blood showed that 65% of the nuclei displayed two signal for the X samples. chromosome and four to five signals for the MAGE1 gene The FISH analysis of tumour 7 (corresponding to a male (Figure 2), thus corroborating the CGH and Southern results.

Leukemia Chromosomal and gene amplifications in T-NHL M Renedo et al 1631 the minimal gained region at Xq26–28. Some cytogenetic studies have shown gains of chromosome X in Tcell lym- phomas, in accordance with our CGH results. Gains of differ- ent regions involving chromosomes 3, 7 and X are frequently observed in other types of B cell lymphoid neoplasms: mar- ginal B cell,14 follicular,15 diffuse large B cell,16 mantle cell, and mediastinal thymic B cell lymphomas,17 suggesting that these regions may include genes associated with lymphoid neoplasms. Other recurrent gains involved the whole or part of chromo- somes 17 (10%), 19 (10%), 20 (10%), and other chromosomal regions as 9q34 (10%) or 11q13 (8%). Thus, CGH studies could provide new information concerning other alterations not found by conventional cytogenetics, such as the gain of 9q34 region, which has rarely been observed in lymphomas. Figure 3 Southern blot analysis of the SH2D1A located in Xq26– The most recurrent deletions were found at 13q (10%), and 28 and the control gene (␤-actin), performed in three tumors with gain of this region (cases 7a, 7b and 25), two lymphomas with CGH normal CGH allowed us to identify a small region commonly deleted profiles of X chromosome (cases 3 and 26) and normal sample of at 13q14–21. Losses of 13q seem to be frequent in these lym- peripheral blood (PB). The sex of the patients is also shown (M, male; phomas. These deletions have also been frequently detected F, female). The expected gene copy number in the three tumors is with CGH in multiple myeloma,18 breast cancer, and less fre- represented below the bands. quently in other lymphoid malignancies.14,19,20 The minimal deleted region suggests the location of important genes that are probably involved in multiple tumors. Discussion In addition, CGH analysis could help to identify changes occurring during the evolution of the disease. Most of the In contrast with B cell NHL, cytogenetic analysis of Tcell NHL cases (four out of seven cases) in which we could study serial reveals only a weak relationship between cytogenetic findings samples, showed an accummulation of aberrations at relapse. and histopathological classification. Moreover, except for Two of them changed from a normal to an abnormal profile, some cytogenetic results, such as the presence of an with gains or losses at the chromosomes more frequently additional X chromosome, structural aberrations of the short involved in these lymphomas such us the X, 17 or 13q. The arm of chromosome 1 and complex karyotypes, associated 13 other two cases maintained the alterations previously detected with a poor prognosis, little is known about the prognostic in these samples, together with the appearance of new alter- impact of chromosome abnormalities in Tcell NHL. Thislack ations of different genetic regions. of information is also evident when it comes to CGH studies, Southern blot analysis has been used to confirm amplifi- while there are different reports about the CGH profile in B- cations of specific genes encoded in chromosomal regions NHL, there are no studies of T-NHL series that have looked showing genomic gains by CGH technique.21 In the present for gains or losses of DNA sequences in the genome of study, gain of Xq26–28 region detected by CGH was con- these lymphomas. firmed by Southern blot and FISH analysis of three genes Due to the scarcity of information about these features, we located on it (SH2D1A, MAGE1, MTCP1), corroborating the performed CGH studies on 37 Tcell lymphomas, in order to CGH results. In addition, gains detected by CGH in this region detect cryptic quantitative genetic alterations, that could corresponded to low level amplifications, and the same gen- otherwise have been masked, and try to establish whether quantitative genetic imbalances could have a clinical value. etic event (two- to three-fold gain) was found with probes cor- In all cases, the immunophenotype and morphology was responding to that particular regions by the Southern blot and compatible with the diagnosis of Tcell lymphomas. To the FISH analysis (Figure 2). characterize these tumors better, the presence of clonal tumo- In general, amplifications or deletions of specific regions ral cells was analyzed by molecular rearrangements of TCR- suggest that important genes located in these chromosomal gamma chain gene. Most of these lymphomas (63%) showed regions are possibly involved in the tumoral pathogenesis of monoclonal rearrangements, a similar percentage to that tumors. In this way, several genes with importance in the found in other Tcell NHLs series. biology and pathology of B-NHL have now been associated We have found a 59% of samples at diagnosis with abnor- with specific regions, such as REL (2p13), BCL-6 (3q27), MYC mal profiles, and a ratio of 66% when the relapse samples (8q24), TEL (12p13) or BCL-2 (18q21). We searched the gen- were included. This is a similar percentage to that reported etic databases for genes located in the most important regions when cytogenetic techniques were used in these lym- with gains or losses in the Tcell lymphomas. TheXq26–28 phomas,3–5 although some cases with normal CGH profiles region contains near 130 known genes and 80 ESTs, some of could in fact be cases with cytogenetic balanced abnormali- them clearly related to cancer. In fact, the genes that we have ties (ie balanced translocations). analyzed have been defined as and are altered in With regard to the possible correlation between abnormal different neoplasms, including some Tcell leukemias. 22 CGH profile and different subtypes of Tcell lymphomas, CGH With regard to clinical presentation, evolution and prog- did not reveal any chromosomal region to be more frequently nosis, cases with normal and abnormal CGH profiles showed associated with morphologically defined subtypes, thus different clinical patterns, whereby a higher incidence of cases reflecting the heterogeneity that characterizes these neo- with advanced disease (stages III–IV), overexpression of P53 plasms. protein and increased proliferative index were observed in The most frequent DNA imbalances of our series were gains cases with altered CGH profiles compared with those with involving the whole or part of the chromosome X (19%), with normal CGH profiles. These differences were statistically sig-

Leukemia Chromosomal and gene amplifications in T-NHL M Renedo et al 1632 nificant and, therefore, CGH analysis could be used as an Benitez J, Oliva H, Delsol G. A multiparametric study of malignant index for diagnosis. lymphoma of mucosa associated lymphoid tissue (Malt). Leuk To date, this is the first CGH study in T cell NHLs. Our Lymphoma 1992; 8: 87–96. 10 Trainor JK, Brisco MJ, Wan JH, Nech S, Grist S, Morley AA. Gene results provide important and previously unknown infor- rearrangement in B- and T-lymphoproliferative disease detected mation about the quantitative genetic abnormalities involved by the polimerase chain reaction. Blood 1991; 78: 192–196. in tumoral pathogenesis of Tcell NHLs. CGH confirmed the 11 Kallionemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, implication of gains of chromosome X in Tcell NHLs, pre- Waldman F, Pinkel D. Comparative genomic hybridization for viously reported from the use of cytogenetic techniques,4,5 molecular cytogenetic analysis of solid tumours. Science 1992; and identified other chromosomes with gains or deletions, 258: 818–821. 12 Kallionemi OP, Kallioniemi A, Piper J, Isola J, Waldman F, Gray allowing us to define more accurately the commonly gained JW, Pinkel D. Optimizing comparative genomic hybridization for regions as bands Xq26–q28, 19q13.1, 9q34, 11q13, 17q12– analysis of DNA sequence copy number changes in solid tumours. q21 or 13q14–q21, although more studies are necessary to Genes Chromosom Cancer 1994; 10: 231–243. clarify the role of the genes located in these regions. 13 Schlegelberger B, Zwingers T, Hohenadel K, Henne-Bruns D, Schmitz N, Haferlach T, Tirier C, Bartels H, Sonnen R, Kuse R et al. Significance of cytogenetic findings for the clinical outcome in patients with T-cell lymphoma of angioimmunoblastic lymphad- Acknowledgements enopathy type. J Clin Oncol 1996; 14: 593–599. 14 Dierlamm J, Rosenberg C, Stul M, Pittaluga S, Wlodarska I, We thank Alicia Barroso and Jesus Gallego for their technical Michaux L, Dehaen M, Vehoef G, Thomas J, de Kelver W, Bakker- assistance. This work has been partially supported by CAM Schut T, Cassiman JJ, Raap AK, De Wolf-Peeters C, Van den Ber- 99/086/0008, and CAM 08.1/0020/00 proyects. A Martinez- ghe H, Hagemeijer A. Characteristic pattern of chromosome gains Ramirez is a fellow of the Instituto de Salud Carlos III. and losses in marginal zone B cell lymphoma detected by com- parative genomic hybridization. Leukemia 1997; 11: 747–758. 15 Avet-Loiseau H, Vigier M, Moreau A, Mellerin MP, Gaillard F, Harousseau JL, Bataille R, Milpied N. Comparative genomic References hybridization detects genomic abnormalities in 80% of follicular lymphomas. Br J Haematol 1997; 97: 119–122. 1 Koeppen H, Vardinam J. New entities, issues and controversies in 16 Monni O, Joensuu H, Franssila K, Knuutila S. DNA copy number the classification of malignant lymphoma. Semin Oncol 1998; 25: changes in diffuse large B-cell lymphoma. Comparative Genomic 421–434. Hybridization study. Blood 1996; 12: 5269–5278. 2 Sandberg AA. The Chromosomes in Human Cancer and Leuke- 17 Knuutila S, Bjo¨rkqvist AM, Autio K, Tarkkanen M, Wolf M, Monni mia. Elsevier: New York, 1990. O, Szymanska J, Larramendy ML, Tapper J, Pere H, EI-Rifai W, 3 Schlegelberger B, Zhang Y, Weber-Matthiesen K, Grote W. Detec- Hemmer S, Wasenius VM, Vidgren V, Zhu Y. DNA copy number tion of aberrant clones in nearly all cases of angioimmunoblastic amplifications in human neoplasms. Am J Pathol 1998; 152: lymphadenopathy with dysproteinemia-type T-cell lymphoma by 1107–1126. combined interphase and metaphase cytogenetics. Blood 1994; 18 Cigudosa JC, Rao PH, Calasanz MJ, Odero MD, Michaeli J, Jhan- 84: 2640–2648. war SC, Chaganti RSK. Characterization of nonrandom chromo- 4 Schlegelberger B, Himmler A, Bartles H, Kuse R, Sterry W, Grote somal gains and losses in multiple myeloma by comparative gen- W. Recurrent chromosome abnormalities in peripheral T-cell lym- omic hybridization. Blood 1998; 8: 3007–3010. phomas. Cancer Genet Cytogenet 1994; 78: 15–22. 19 Avet-Loiseau H, Andree-Ashley LE, Moore D 2nd, Mellerin MP, 5 Schlegelberger B, Feller AC. Classification of peripheral T-cell Feusner J, Bataille R, Pallavicini MG. Molecular cytogenetic lymphomas: cytogenetic findings support the updated Kiel classi- abnormalities in multiple myeloma and plasma cell leukemia fication. Leuk Lymphoma 1996; 20: 411–416. measured using comparative genomic hybridization. Genes Chro- 6 Weidmann E. Hepatosplenic T-cell lymphoma. A review on 45 mosom Cancer 1997; 19: 124–133. cases since the first report describing the disease as a distinct lym- 20 Barth T, Do¨hner H, Werner CA, Stilgenbauer S, Schlotter M, Pawl- phoma entity in 1990. Leukemia 2000; 14: 991–997. ita M, Lichter P, Mo¨ller P, Bentz M. Characteristic pattern of 7 Harris NL, Jaffe ES, Diebold J, Flandrin G, Muller-Hermelink HK, chromosomal gains and losses in primary large B-cell lymphomas Vardiman J, Lister TA, Bloomfield CD. The World Health Organi- of the gastrointestinal tract. Blood 1998; 91: 4321–4330. zation classification of neoplastic diseases of the haematopoietic 21 Rao PH, Houldsworth J, Dyomina K, Parsa NZ, Cigudosa JC, Louie and lymphoid tissues: report of the Clinical Advisory Committee DC, Popplewell L, Offit K, Jhanwar SC, Chaganti RS. Chromo- meeting-Airlie House, Virginia, November 1997. J Clin Oncol somal and gene amplification in diffuse large B-cell lymphoma. 1999; 17: 3835–3849. Blood 1998; 92: 234–240. 8 Stansfeld AG, Diebold J, Kapanci Y, Kele´nyi G, Lennert K, 22 Stern MH, Soulier J, Rozenzwajg M, Nakahara K, Canki-Klain N, Mioduszewska O, Noel H, Rilke F, Sundstrom C, Van Unnik JAM, Aurias A, Sigaux F, Kirsch IR. MTCP1: a novel gene on the human Wright DH. Updated Kiel classiffication for Lymphomas (letter). chromosome Xq28 translocated to the Tcell receptor alpha/delta Lancet 1988; 1: 292–293. locus in mature Tcell proliferations. 1993; 8: 2475– 9 Rivas C, Echezarreta G, Garcia R, Santos M, Santon A, Robledo M, 2483.

Leukemia