Prevalence and clinical implications of bone marrow ... - Nature

0 downloads 0 Views 165KB Size Report
Jul 28, 2005 - Anaplastic large cell lymphoma (ALCL) harbors the reciprocal chromosomal translocation t(2;5)(p23;q35) in approximately. 80% of the cases.
Leukemia (2005) 19, 1643–1647 & 2005 Nature Publishing Group All rights reserved 0887-6924/05 $30.00 www.nature.com/leu

Prevalence and clinical implications of bone marrow involvement in pediatric anaplastic large cell lymphoma L Mussolin1, M Pillon1, ES d’Amore2, N Santoro3, A Lombardi4, F Fagioli5, L Zanesco1 and A Rosolen1 1 Clinica di Oncoematologia Pediatrica, Azienda Ospedaliera-Universita` di Padova, Padova, Italy; 2U.O. Anatomia Patologica, Ospedale S. Bortolo, Vicenza, Italy; 3U.O. Pediatria, Azienda Ospedaliera-Policlinico, Bari, Italy; 4Clinica di Oncoematologia Pediatrica, Ospedale Bambino Gesu`, Rome, Italy; and 5OIRM S. Anna, Torino, Italy

Anaplastic large cell lymphoma (ALCL) harbors the reciprocal chromosomal translocation t(2;5)(p23;q35) in approximately 80% of the cases. The genes involved are nucleophosmin (NPM) and anaplastic lymphoma kinase (ALK) and the resulting chimeric NPM–ALK protein is thought to play a key role in the pathogenesis of t(2;5) positive ALCL. Few data on bone marrow (BM) involvement in ALCL have been published and they mostly rely on morphological examination of BM smears. We studied 52 ALCL for NPM–ALK expression by RT-PCR: 47/52 biopsies were positive. In 41 of the 47 cases we obtained the BM at diagnosis and investigated the prevalence of minimal BM infiltration by RT-PCR and real-time PCR. Minimal disseminated disease was positive in 25/41 patients (61%), of whom six had morphologically infiltrated BM. Survival analysis demonstrated a 5-year progression-free survival of 41711% for patients with molecularly positive BM vs 100% for patients with negative BM (P ¼ 0.001). These results suggest that minimal BM involvement at diagnosis is a common event in pediatric ALCL and that minimal BM disease monitoring could identify patients at risk of relapse. Leukemia (2005) 19, 1643–1647. doi:10.1038/sj.leu.2403888; published online 28 July 2005 Keywords: ALCL; minimal disseminated disease; NPM–ALK; RT-PCR

Introduction Anaplastic large cell lymphoma (ALCL) belongs to the group of high-grade non-Hodgkin’s lymphomas (NHLs) and typically presents as an aggressive systemic disease, with or without extranodal involvement in children and young adults. ALCL cells are characterized by the expression of the CD30/Ki-1 and more often by a T-cell phenotype, although in a minority of cases neither T-cell markers nor T-cell receptor rearrangements can be detected.1,2 ALCL shows a broad morphological spectrum, ranging from cases with prevalence of small monomorphic tumor cells to cases characterized by large anaplastic tumor cells, as well as a combination of both features.3 ALCL is frequently associated with the t(2;5)(p23;q35) chromosomal translocation, which gives rise to the fusion gene NPM–ALK.4,5 Nucleophosmin (NPM) is a widely expressed gene responsible for shuttling proteins between the nucleus and cytoplasm. Anaplastic lymphoma kinase (ALK) is a leukocyte tyrosine kinase, usually absent in normal lymphoid tissues.6,7 The NPM–ALK fusion gene encodes an 80 kDa chimeric protein that is thought to play a key role in lymphomagenesis through aberrant phoshorylation of intracellular substrates.8,9 Correspondence: Dr A Rosolen, Clinica di Oncoematologia Pediatrica, Azienda Ospedaliera-Universita` di Padova, Via Giustiniani 3, Padova 35128, Padova, Italy; Fax þ 39 049 8213510; E-mail: [email protected] Received 18 March 2005; accepted 7 June 2005; published online 28 July 2005

Recent data suggest that t(2;5) positive ALCL respond better to therapy and have a higher survival rate than translocation negative ALCL, thus implying that NPM–ALK expression may have a significant prognostic impact.10 In children, however, the prognostic significance of the translocation is still uncertain. Bone marrow (BM) involvement in ALCL is considered an uncommon event.11 This assumption may be influenced by the subtle nature of the BM involvement and by the difficulty of its detection based on routine morphologic examination. Since the chimeric NPM–ALK transcript, encoded by the derivative chromosome 5, can be detected by reverse transcriptase polymerase chain reaction (RT-PCR) using NPM- and ALKspecific primers, we conducted a prospective study to evaluate the prevalence of minimal BM infiltration at diagnosis (minimal disseminated disease-MDD) by qualitative RT-PCR and quantitative real-time RT-PCR in a series of children affected by ALCL treated in the Italian Association of Pediatric Hematology and Oncology (AIEOP) centers. In addition, we analyzed the BM status in relation to specific clinical features and performed a comparative study between RT-PCR results and BM biopsy and BM smear microscopic examinations.

Materials and methods

Patients and specimens Excess fresh tumor tissue from diagnostic biopsies of children affected by ALCL enrolled between October 1997 and December 2003 in AIEOP trials based on BFM-like strategies12 for pediatric NHL was selected by the local pathologist, maintained in culture medium and received in our laboratory within 24 h from biopsy. Tissue was processed immediately or frozen and stored at 801C until used. As per protocol, a BM aspirate in sodium citrate was obtained from each patient at diagnosis and from each BM sample, nucleated cells were obtained by differential lysis, whereas tumor biopsies were minced and subsequently mechanically homogenized for RNA extraction. The clinical and research studies were approved by the local ethics committees and informed consent was obtained from all the patients.

Staging Patients were staged according to the St Jude’s classification.13 Each patient was also classified according to two different risk groups: high risk (HR) and low risk (LR). This was based on the result of a multivariate analysis on a large cohort of ALCL patients reported by the French SFOP group14 that demonstrated a significant negative prognostic impact of mediastinal localization or any visceral involvement (liver, spleen or lung) or LDH X800 IU/l, thus featuring an HR group, compared to

MDD in pediatric ALCL L Mussolin et al

1644 children without such features (identified as a LR group). In our study we adopted the same risk definition with the sole exception that LDH cutoff value was twice the institution normal value (n.v.). In this risk classification, BM involvement, as determined by morphological and immunophenotypical analysis, did not bear any impact on prognosis.

Histological and immunohistochemical studies In all the cases histological and immunohistochemical diagnoses were centrally reviewed. Formalin-fixed paraffin-embedded tumor biopsies from all the ALCL cases were analyzed by immunohistochemistry using a wide panel of antibodies including those for T- and B-lineage markers (CD2, CD3, CD4, CD5, CD7, CD8, CD20, CD43, CD45RO, CD79a), NK markers (CD56, CD57), Alk-1, EMA, clusterin and cytotoxic proteins (Tia1, Granzyme B and Perforin). BM biopsy performed at the initial staging, was fixed in 10% formalin with Na2POH4 and NAH2PO4, subjected to decalcification using Mielodec (Bio-optica, Milan, Italy) and paraffin embedded. For each bone marrow biopsy 10 hematoxylin-andeosin-stained sections were examined and immunostaining for CD30, CD20, CD3, EMA and ALK1 was also performed.

RT-PCR Total RNA was isolated using RNAzol (Tel-Test Friendswood, USA), following the manufacturer’s instructions. An amount of 1 mg of total RNA was reverse transcribed using SuperScript II reverse transcriptase (Life Technologies, Milan, Italy) and random hexamers. For each sample, b2-microglobulin expression was assessed as a control for the presence of amplifiable RNA and the efficiency of reverse transcription. The 50 and 30 primers specific for the chimeric transcript NPM–ALK were TCCCTTGGGGGCTTTGAAATAACACC (50 NPM) and CGAGG TGCGGAGCTTGCTCAGC (30 ALK). Each reaction mixture contained 10  buffer, 1.5 mM MgCl2, 1.6 mM dNTPs, 400 nM of each primer, 0.2 IU of Taq polymerase and 5% of the RT product in a final 20 ml reaction volume. PCR reaction consisted of initial denaturation at 941C for 2 min, followed by 40 cycles of 941C for 15 s, 681C for 15 s, 721C for 30 s and a final extension at 721C for 10 min. PCR products were analyzed by 3% agarose gel electrophoresis and visualized under UV illumination after ethidium bromide staining. Ladder 50 (Invitrogen, Milan, Italy) was used as a molecular weight standard.

Real-time PCR The expression levels of 11 candidate endogenous control genes were assessed by real-time quantitative RT-PCR using the 50 nuclease technology and the Human TaqMans predeveloped assay reagents (PDARs) endogenous controls (ABI, Foster City, CA, USA). Five of 20 ml of the RT reaction was used as a template in a 50 ml reaction for the real-time detection of NPM–ALK fusion transcript using the TaqMan Universal Master Mix and the ABI Prism 7000 sequence detection system. The PCR mixtures contained Master Mix 1  , 900 nM of each forward and reverse primer and 200 nM of Taqman probe. Primer directed to the ALK portion of the transcript was 50 -TCTGCATGGCTTGCAGCTC Leukemia

and the forward NPM primer was 50 -GGGCCAGTGCATAT TAGTGGA. The TaqMan probe was 6-carboxyfluorescein (FAM) labeled and bridged the NPM–ALK breakpoint (AGCACTTAG TAGTGTACCGCCGGAAGCACC). The quality of the newly synthesized cDNA was verified during the same assay by amplification of the best housekeeping gene identified in the preliminary test. Amplification conditions were 2 min at 501C and 10 min at 951C, followed by 45 cycles of denaturation (951C, 15 s) and extension (601C, 1 min). Results of real-time PCR were expressed as Ct value, which represents the cycle at which fluorescence raised above a threshold value, by using a standard curve to determine the starting target quantity. Each sample was normalized by dividing the NPM–ALK target quantity by the housekeeping target amount. The standard curves for both NPM–ALK and the housekeeping gene were obtained with 5-fold serial dilutions of cDNA in deionized water prepared from total RNA extracted from the Karpas-299 cell line. Each PCR run was performed in triplicate and the mean values of results were obtained.

Statistical analysis Overall survival (OS) and progression-free survival (PFS) were analyzed by the Kaplan–Meier method15 on patients with at least 12 months follow-up from stop therapy (15 months from diagnosis). OS was defined as the time from diagnosis to death due to any cause or to last follow-up. PFS was calculated from the date of diagnosis to relapse, disease progression after partial remission (PR) or to the date of last follow-up. The log-rank test was used for univariate analysis16 with the inclusion into the Cox multivariate model of variables with a P-value o0.1.17 The association of the MDD with specific clinical characteristics (stage, mediastinal or visceral involvement and LDH 42 n.v.) was analyzed by the w2 test or Fisher’s exact test. All P-values are two-sided, with a type I error rate fixed at 0.05. Statistical analysis was carried out using the SAS statistical program (SAS-PC, version 8.2; SAS Institute Inc., Cary, NC, USA). The cutoff limit for the current study was the end of February 2005.

Results

Patients A total of 52 patients with ALCL, 17 females and 35 males, aged 1.4–17.4 years (median age, 9.8 years) were enrolled in this study. Central review of histological diagnosis demonstrated that most cases were common type ALCL (39/52), characterized by the presence of abundant hallmark cells, three were small cell, three lymphohistiocytic, two giant cell and five mixed morphology subtypes. As to the phenotype, 38/52 cases were of T-cell lineage, ten patients lacked T-cell antigen expression (null ALCL) and four were undetermined due to the incomplete series of T-cell markers studied. All cases were CD30 and EMA positive, whereas 47 out of 52 (90%) reacted with the Alk-1 antibody.

RT-PCR analysis of BM and relationship to clinical features With the aim of studying MDD in ALCL, we ascertained whether RT-PCR for the NPM–ALK chimeric gene transcript had

MDD in pediatric ALCL L Mussolin et al

sufficient sensitivity. Increasing dilutions of the translocation positive Karpas-299 ALCL cells in Jurkat T-cell acute leukemia cells showed a sensitivity of 106 (Figure 1a). Tumor specimens from all the 52 ALCL patients were assayed for NPM–ALK transcript expression by RT-PCR. NPM–ALK was detected in 47/52 (90%) tumors, confirming the immunohistochemical findings. BM at diagnosis was available in 41/47 positive cases and in 25 of them (61%) NPM–ALK transcript was detected (Figure 1b). The main clinical and molecular characteristics of the 41 patients are listed in Supplementary Information. Only 6/41 (15%) BM aspirates were positive at standard microscopic and immunophenotypic analyses. In 35/41 patients, BM biopsy was also performed at initial staging: six were positive. In four of them, malignant cells were also detected in BM smears. Five out of six patients with positive BM biopsy had an RT-PCR positive BM, whereas of the 29 negative BM biopsies, 14 were RT-PCR positive. The distribution of patients according to risk groups was 22/41 in HR (54%) and 19/41 in LR (46%). A total of 17/22 HR patients (77%) had an RT-PCR positive BM, whereas BM positivity was detected in 8/19 LR patients (42%) (P ¼ 0.02; Fisher’s exact test). Significant association was found between MDD positivity and presence or absence of mediastinal involvement (13/16 vs 12/25, respectively, w2 ¼ 0.03), whereas no association was observed with LDH X or o2 n.v. (8/10 vs 17/31; Fisher’s exact test ¼ 0.3), presence or absence of visceral involvement (7/9 vs 18/32, Fisher’s exact test ¼ 0.4) and localized (stage I–II) vs highstage disease (stage III–IV) (3/6 vs 22/35, Fisher’s exact test ¼ 0.7).

Analysis of BM involvement by real-time PCR

1645

To identify the most appropriate endogenous control gene for the quantification of NPM–ALK expression in BM of ALCL patients, we conducted a preliminary expression analysis of 11 housekeeping genes in BM aspirates of three patients at diagnosis. In our samples GAPDH was the best candidate control gene because its expression showed the lowest variability across the test samples. To assess the specificity of the method, NPM–ALK expression was measured in BL41 and CA46 Burkitt’s lymphoma cell lines where we could not detect any signal. Of the 25 RT-PCR positive BM samples, we could study NPM–ALK expression by real-time PCR in 20 (Figure 1c). All of them were positive for NPM–ALK. Ct values ranged from 22 to 38. In one case there was a strong positivity (Ct ¼ 22); the other patients had an intermediate or a weak positivity. The expression level (measured as NPM–ALK/GAPDH quantity ratio) ranged from 0.0007 to 1.6 with a median of 0.01. We also studied nine patients with a negative BM by standard RT-PCR and obtained a barely detectable signal in two of them (Ct ¼ 40).

Survival analysis Survival analysis was conducted on 35 children. Six out of 41 patients (No. 36–41, Supplementary Information) with NPM–ALK positive tumor and available BM were excluded because of short follow-up (o15 months). CRs were achieved in 34 out of 35 patients (97%), whereas one patient had a

Figure 1 RT-PCR detection of NPM–ALK transcript. (a) RT-PCR detection of NPM–ALK in increasing dilutions of NPM–ALK positive Karpas299 ALCL cells in leukemic blasts showed a sensitivity of 106. (b) NPM–ALK expression in tumor (T) and bone marrow (BM) samples at diagnosis and in negative (CA46) and positive (Karpas) control cells. (c) NPM–ALK mRNA expression by quantitative real-time PCR in a positive (01) and negative (02) BM. Leukemia

MDD in pediatric ALCL L Mussolin et al

1646 relapse. After a median observation time of 3.1 years (range 1.3– 7.2 years), the 5-year estimated PFS was 41711% in MDDpositive patients compared to 100% in MDD-negative patients (P ¼ 0.001) (Figure 2a). The OS7standard error was 9575% for MDD-positive and 100% for MDD-negative patients (Figure 2b). Survival analysis performed on the whole patient population (41 children) gave same results (PFS 43711% vs 100%, P ¼ 0.001; OS 9575% vs 100%, P ¼ 0.4). The results of the PFS univariate analysis demonstrated that MDD was the only significant factor on survival among those considered (Table 1). The risk group, defined by disease localization and LDH value did not reach a statistical significance by univariate analysis in our series (50712% vs 80710% for HR and LR, respectively; P ¼ 0.08).

Discussion

Figure 2 Survival analysis in pediatric ALCL patients according to MDD status as determined by NPM–ALK expression in BM at diagnosis by RT-PCR. (a) Progression-free survival; (b) overall survival. BM negative (broken line); BM positive (solid line).

Table 1 PFS univariate analysis. Patient characteristics and outcome: prognostic value of the clinical features Characteristics

No. of patients

No. of events

PFS % (E.S.)

P-value

MDD Negative Positive

13 22

0 12

100 41 (11)

0.001

Stage IFII III–IV

6 29

2 10

67 (19) 63 (10)

0.8

LDH o2  N X2  N

28 7

9 3

66 (9) 54 (20)

0.5

Mediastinum No Yes

20 15

5 7

73 (10) 55 (13)

0.2

Visceral involvement No Yes

28 7

9 3

66 (9) 54 (20)

0.6

progressive disease after a partial CR and reached a late CR with a second-line treatment including BM stem cell rescue. There were 11 relapses: 10 of the patients are still alive in second CR obtained with second line chemotherapy alone (N ¼ 5) or with intensified treatment followed by BM stem cell rescue (N ¼ 5), while one patient died of disease progression after the third Leukemia

BM involvement in ALCL is considered an uncommon event. 11,18 However, results of the few studies published on this issue are based on conventional analyses characterized by a low sensitivity and this aspect is particularly critical in ALCL whose cells are difficult to identify in BM based on morphology alone. To overcome this limitation, we used an RT-PCR assay for the identification of NPM–ALK transcript that, in our hands, could detect 106 tumor cells. By using this technique, we established that MDD had an unsuspected high prevalence in pediatric ALCL with 61% of the patients showing an NPM–ALK positive BM at diagnosis. Only 15% were positive at the standard microscopic and immunophenotypic analysis and 17% of children had a positive BM biopsy, thus confirming the significantly higher sensitivity of the RT-PCR assay compared to standard microscopic examinations. In one patient, RT-PCR was also critical in excluding a false positivity based on the sole morphological examination of BM biopsy (patient No. 33). Given the high prevalence of MDD at diagnosis we wanted to quantify the tumor infiltration in BM. Thus, we established a real-time RT-PCR assay for the quantification of NPM–ALK transcript expression in BM specimens, as was previously performed for other molecular targets in different diseases.19–21 Interestingly, as reported in other hematologic malignancies,22 we found a considerable heterogeneity of NPM–ALK expression levels in patients whose BM was positive by RT-PCR. Whether this bears any biological or clinical meaning needs to be clarified. In this context though, it would be of interest to assess whether the NPM–ALK expression levels at diagnosis correlate with BM infiltration response kinetics and whether expression levels at different time points during chemotherapy may predict the risk of disease progression in patients with morphological remission. When NPM–ALK expression was quantified relative to the GAPDH housekeeping gene, the NPM–ALK/GAPDH ratio ranged from 0.0007 to 1.6. We did not find a correlation between NPM–ALK expression levels and risk of relapse. This may be due to the relatively low number of patients enrolled in the study, and may also suggest that within the MDD positive cohort of patients, levels of MDD may not further distinguish children with differential risk of failure. Together with the assessment of MDD prevalence in ALCL patients, the main aim of our study was to determine whether any correlation between BM status at diagnosis and clinical features existed in ALCL. We observed a significant association between BM tumor involvement and the Risk Group. Overall 77% of the patients with ‘high risk’ features had positive BM by RT-PCR compared to only 42% in the ‘low risk’ group.

MDD in pediatric ALCL L Mussolin et al

1647 The significant association between mediastinal involvement and BM positivity raises the possibility of monitoring ALCL patients for risk of relapse both by imaging studies and by MRD analysis. Interestingly, none of the four patients with HR features but with molecularly negative BM had a recurrence. Similarly, none of the LR patients with RT-PCR negative BM experienced a relapse. Taken together these findings would suggest that, independently of the clinical risk group definition, absence of BM involvement at diagnosis would identify a subgroup of children with good prognosis. By univariate analysis a significant lower PFS, but not OS, was demonstrated in patients with MDD compared to the negative counterpart. A difference in OS was reported in a previous study, including adult and pediatric ALCL patients treated, however, with different therapy approaches, where BM involvement was defined on the basis of standard morphological and immunohistochemical analysis.11 Despite recent improvements in outcome, ALCL has a noticeable risk of relapse and, although most children can be salvaged by second or further line therapy, early relapse still represents an unfavorable event that implies aggressive therapies and possibly a dismal outcome. The identification of a risk factor such as MDD that might significantly contribute to predict outcome when evaluated by multivariate analysis on large series of patients, might enable us to design more efficacious treatments for the cohort of patients who risk to succumb to their disease. This may possibly include a better definition of those children who might benefit from high dose-intensity therapy followed by BM rescue. Alternatively, one may plan treatment reduction for the category of patients with a very low risk of failure (ie, children without BM involvement). In conclusion, our data demonstrate for the first time that BM is frequently involved in children with ALCL at diagnosis and that MDD is associated with a high risk of relapse. However, larger studies are needed to draw definitive conclusions about a possible independent prognostic value of MDD in ALCL.

4

5

6

7

8

9 10 11 12

13 14

Acknowledgements This work was supported by Fondazione Citta` della Speranza, by a grant from MIUR-CNR and by AIL (Associazione Italiana contro le Leucemie). L Mussolin is a fellow of Fondazione Citta` della Speranza. We thank Gloria Tridello for the statistical analysis, Ilaria Zecchini for data collection and management and Angelica Zin for technical support. We are grateful to the clinicians of the Italian Association of Pediatric Hematology and Oncology (AIEOP) for providing biological samples and patient information.

15 16 17 18 19

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

20

References 1 Falini B, Pileri S, Pizzolo G, Durkop H, Flenghi L, Stirpe F et al. CD30 (Ki-1) molecole: a new cytokine receptor of the tumor necrosis factor receptor superfamily as a tool for diagnosis and immunotherapy. Blood 1995; 85: 1–14. 2 Stansfeld AG, Diebold J, Kapanci Y, Rilke F, Kelenyi G, Sundstrom C et al. Updated Kiel classification for lymphomas. Lancet 1988; I: 292–293. 3 Benharroch D, Bedoyan Z, Lamant L, Amin C, Brugieres L, Terrier-Lacombe MJ et al. ALK-positive lymphoma: a single

21

22

disease with a broad spectrum of morphology. Blood 1998; 91: 2076–2084. Rimokh R, Magaud JP, Berger F, Samarut J, Coiffier B, Germain D et al. A translocation involving a specific breakpoint (q35) on chromosome 5 is characteristic of anaplastic large cell lymphoma (‘Ki-1 lymphoma’). Br J Haematol 1989; 71: 31–36. Lamant L, Meggetto F, Saati T, Brugieres L, de Paillerets BB, Dastague N et al. High incidence of the t(2;5)(p23;q35) translocation in anaplastic large cell lymphoma and its lack of detection in Hodgkin’s disease. Comparison of cytogenetic analysis, reverse transcriptase-polymerase chain reaction, and P-80 immunostaining. Blood 1996; 87: 284–291. Morris SW, Naeve C, Mathew P, James PL, Kirstein MN, Cui X et al. ALK, the chromosome 2 gene locus altered by the t(2;5) in non-Hodgkin’s lymphoma, encodes a novel neural receptor tyrosine kinase that is highly related to leukocyte tyrosine kinase (LTK). Oncogene 1997; 14: 2175–2188. Iwahara T, Fujimoto J, Wen D, Cupples R, Bucay N, Arakawa T et al. Molecular characterization of ALK, a receptor tyrosine kinase expressed specifically in the nervous system. Oncogene 1997; 14: 439–449. Fujimoto J, Shiota M, Iwahara T, Seki N, Satoh H, Mori S et al. Characterization of the transforming activity of p80, a hyperphosphorilated protein in a Ki-1 lymphoma cell line with chromosomal translocation t(2;5). Proc Natl Acad Sci USA 1996; 93: 4181–4186. Kuefer MU, Look AT, Pulford K, Behm FG, Pattengale PK, Mason DY et al. Retrovirus-mediated gene transfer of NPM–ALK causes lymphoid malignancy in mice. Blood 1997; 90: 2901–2910. Falini B, Pileri S, Zanzini PL, Carbone A, Zagonel V, Wolf-Peeters C et al. ALK+ lymphoma: clinico-pathological findings and outcome. Blood 1999; 93: 2697–2706. Fraga M, Brousset P, Schlaifer D, Payen C, Robert A, Rubie H et al. Bone marrow involvement in anaplastic large cell lymphoma. Am J Clin Pathol 1995; 103: 82–89. Seidemann K, Tiemann M, Schrappe M, Yakisan E, Simonitsch I, Janka-Schaub G et al. Short-pulse B-non-Hodgkin lymphoma-type chemotherapy is efficacious treatment for pediatric anaplastic large cell lymphoma: a report of the Berlin–Frankfurt–Munster Group Trial NHL-BFM 90. Blood 2001; 97: 3699–3706. Murphy SB. Classification, staging and end results of treatment of childhood non-Hodgkin’s lymphomas: dissimilarities from lymphomas in adults. Semin Oncol 1980; 7: 332–339. Brugie`res L, Le Deley MC, Pacquement H, Meuerian-Bedoyan Z, Terrier-Lacombe MJ, Robert A et al. CD 30+ anaplastic large-cell lymphoma in children: analysis of 82 patients enrolled in two consecutive studies of the French Society of Pediatric Oncology. Blood 1998; 92: 3591–3598. Kaplan EL, Meier P. Non parametric estimation from incomplete observation. J Am Stat Assoc 1958; 53: 457–481. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J R Stat Soc A 1972; 135: 185–198. Cox DR. Regression models and life tables. J R Stat Soc 1972; 34: 187–220. Stein H, Dallenbach F. Diffuse large cell lymphomas of B and T cell type. In: Knowles DM (ed). Neoplastic Hematopathology. Williams and Wilkins: Baltimore, 1992, pp 675–714. Gerard CJ, Olsson K, Ramanathan R, Reading C, Hanania EG. Improved quantitation of minimal residual disease in multiple myeloma using real-time polymerase chain reaction and plasmidDNA complementary determining region III standards. Cancer Res 1998; 58: 3957–3964. Marcucci G, Livak KJ, Bi W, Strout MP, Bloomfield CD, Caligiuri MA et al. Detection of minimal residual disease in patients with AML1/ETO-associated acute myeloid leukemia using a novel quantitative reverse transcription polymerase chain reaction assay. Leukemia 1998; 12: 1482–1489. Mensink E, Van de Locht A, Schattenberg A, Linders E, Schaap N, Geurts van Kessel A et al. Quantitation of minimal residual disease in Philadelphia chromosome positive chronic myeloid leukaemia using real-time quantitative RT-PCR. Br J of Haematol 1998; 102: 768–774. Viehmann S, Teigler-Schlegel A, Bruch J, Langebrake C, Reinhardt D, Harbott J. Monitoring of minimal residual disease (MRD) by real-time quantitative reverse transcription PCR (RQ-RT-PCR) in childhood acute myeloid leukaemia with AML1/ETO rearrangement. Leukemia 2003; 17: 1130–1136. Leukemia