Bone Marrow Transplantation, (1999) 23, 437–442  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Autologous stem cell transplantation for T and null cell CD30-positive anaplastic large cell : analysis of 64 adult and paediatric cases reported to the European Group for Blood and Marrow Transplantation (EBMT)

R Fanin1, MC Ruiz de Elvira2, A Sperotto1, M Baccarani1 and A Goldstone2 for the EBMT Lymphoma Working Party

1Division of Hematology and Department of Bone Marrow Transplantation, University Hospital, Udine, Italy; and 2Department of Hematology, University College London Hospitals, London, UK

Summary: of the cases, with some recent series showing up to 90% of complete remissions.1–8 Relapse rate ranges widely from Anaplastic large cell lymphoma (ALCL) is a hetero- 25 to 45%, with less than 50% of patients alive and free geneous family of lymphoid tumours, among which the of disease at a median observation time from diagnosis no T and null cell types were recently listed in the REAL longer than 36 months.1–8 The different behaviour in classification as a distinct entity. Reports on autologous relapse rate may reflect the existence of a heterogeneous stem cell transplantation (ASCT) in this group are only family of diseases: ALCL T and null cell common type, occasional. Sixty-four patients with T and null cell ALCL T and null cell Hodgkin-like and diffuse anaplastic ALCL from 25 European centres had been registered large B cell lymphoma.9,10 These three major entities prob- with the European Group for Blood and Marrow ably have different prognosis and outcome, but this is not Transplantation (EBMT) at the onset of this study. The clearly underlined in the papers already published on this median age was 25 years (range 3.2–53.0). Thirty of the subject.11,12 64 patients (47%) were in complete remission (CR), 18 It is surprising that although the clinical, histological and (28%) in partial remission (PR), and the remaining 16 immunophenotypic characteristics indicate that this lym- (25%) had a more advanced or chemotherapy-refrac- phoma may be a group of aggressive lymphoid tumours, tory disease at transplant. Eighty-one percent of the reports on autologous stem cell transplantation (ASCT) in patients were conditioned with chemotherapy alone and ALCL are occasional, and mainly refer to cases in 75% received marrow stem cells. All the patients trans- relapse.1,2,4,5 It was recently shown, in a study performed planted in first CR (15), except one, maintained the CR on a small monocentric series, that primary systemic CD30- over time; six of 15 transplanted in CR subsequent to positive ALCL of the adults is a curable disease when first, six of 18 transplanted in PR and 14 of 16 trans- ASCT is performed as part of first-line treatment.13,14 In planted in refractory or relapsed disease progressed. this study patients with B cell ALCL were included. Actuarial overall survival (OS) at 10 years is 70%. The REAL classification defines T and null cell ALCL Multivariate analysis showed that good status at trans- as a distinct entity with a fairly good prognosis.10 To under- plant, younger age, absence of B symptoms and absence stand the relevance that the inclusion of this lymphoma had of extranodal disease indicated a better prognosis. These for the good results obtained in the monocentric study men- data suggest that ASCT should be considered as a pos- tioned above, we have studied 64 patients affected by T sible treatment for chemosensitive patients in CR or PR. and null cell ALCL autotransplanted and registered with However, definitive conclusions cannot be drawn from the European Group for Blood and Marrow Transplantation this study and a prospective randomised trial between (EBMT) between 1983 and 1996. ASCT and conventional chemotherapy may be indicated. Keywords: anaplastic large cell lymphoma; autologous Patients and methods stem cell transplantation EBMT centres were asked to submit information on all patients with NHL who had been reclassified or newly classified, using the REAL criteria, as T and null cell ALCL The therapy of primary systemic CD30-positive anaplastic and who had been subjected to autologous transplantation large cell lymphoma (ALCL) with third generation regi- between 1983 and 1996. No attempt was made to review mens of chemotherapy (CHT) is successful in about 70% the histology centrally. All centres which participated in the study are contributors of lymphoma cases to the EBMT Correspondence: Dr R Fanin, Division of Hematology, University Hospi- and possess reliable histopathology laboratories (see tal, Ple S Maria della Misericordia, 33100 Udine, Italy Acknowledgements). Received and accepted 5 October 1998 The information requested was: sex, age, histology, ASCT for T and null cell CD30-positive ALCL R Fanin et al 438 phenotype, performance status (ECOG scale),15 B symp- Table 1 Characteristics of the patients at diagnosis toms, nodal involvement, bulk, number and sites of extranodal disease, Ann Arbor stage, LDH, chemotherapy Number % program, radiotherapy, status at transplant, months from diagnosis to ASCT and from ASCT to present, stem cell Patients 64 Sex source, conditioning regimen, response at 90 days, relapse Male 39 60.9 or progression after ASCT, current status, date of death and Female 25 39.1 date of progression/relapse if applicable. Age median (range) 25.2 (3.2–53.0) Disease status was defined as follows: complete Stage remission (CR) as no evidence of disease after treatment; I 3 4.7 partial remission (PR) as less than 50% residual disease as II 18 28.1 compared to pre-treatment. Refractory disease was defined III 25 39.1 IV 16 25.0 as no effect of treatment or a decrease in disease load fol- Unknown 2 3.1 lowing treatment of less than 50%. Responding relapse B symptoms refers to patients with relapsed disease whose disease has No 35 54.7 Yes 29 45.3 responded to salvage chemotherapy prior to transplant. BM involvement No 60 93.7 Yes 3 4.7 Statistics Unknown 1 1.6 Lymph nodes inv above diaphragm Overall survival (OS) was calculated from the date of trans- No 27 42.2 plantation to death or last date seen. Progression-free sur- Yes 36 56.2 vival (PFS) was calculated from date of transplantation to Unknown 1 1.6 Lymph nodes inv below diaphragm date of progression if the patient progressed, or to date of No 28 43.8 death or last date seen if the patient did not progress. The Yes 36 56.2 Kaplan and Meier method was employed to estimate sur- Med involvement 16 No 26 40.6 vival; the log-rank test to detect significant differences. Yes 37 57.8 Multivariate analysis was performed on possible prognostic Unknown 1 1.6 factors using the Cox model.17 Proportionality of hazards Lungs involvement was tested using standard graphical methods. The age- No 57 89.1 Yes 6 9.3 adjusted international prognosis index (AAIPI) was calcu- Unknown 1 1.6 lated for all patients for which the information was avail- Size largest mass able.18 Ͻ5 cm 10 15.6 5–10 cm 26 40.6 Ͼ10 cm 15 23.5 Unknown 13 20.3 Results LDH Normal 29 45.3 High 21 32.8 A total of 64 patients referred from 25 European centres Unknown 14 21.9 was included in the study. Patient details regarding charac- Number extranodal sites teristics at diagnosis and response to first-line treatment are 0 39 61.0 1 8 12.5 reported in Table 1. 2 9 14.0 3–4 2 3.1 Unknown 6 9.4 Clinical features ECOG scale PS 0 23 36.0 Median age at diagnosis was 25 years with 18 patients 1 15 23.4 younger than 20 years. Sixty-four percent of the patients 2 8 12.5 presented with advanced stage disease (III + IV). In about 3 3 4.7 Unknown 15 23.4 60% of them a mediastinal involvement (with bulky mass AAIPI in about 25% of the cases) was documented; as expected, 0, low 9 14.0 bone marrow involvement was occasional. The AAIPI was 1, Low–intermediate 22 34.4 2, High–intermediate 9 14.0 calculated for 70% of the patients, with more than half of 3, High 5 7.8 these patients in the low risk category (AAIPI 0 and 1). Unknown 19 30.0 Sixty-four percent of patients below 20 were in the high Radiotherapy 1st line risk category (AAIPI 2 and 3) compared to only 19% in No 58 90.6 Yes 3 4.7 older patients (Fischer exact test, P Ͻ 0.01). Unknown 3 4.7 Response 1st line CR 32 50.0 First line treatment and status at transplant PR 18 28.1 No response 7 11.0 First-line therapy was based on chemotherapy alone in Progression 1 1.6 more than 90% of the cases with only three patients being Unknown 6 9.3 irradiated before ASCT and none after. A response was obtained or maintained in more than 75% of the patients ASCT for T and null cell CD30-positive ALCL R Fanin et al 439 Table 2 Characteristics at transplant a 100

Number % 80 Patients 64 Age median (range) 25.6 (4.8–54.1) 60 CR at transplant No 34 53.1 Yes (of which 15 on CR1) 30 46.9 40 Months between diagnosis and transplant

median (range) 9.6 (3.6–94.6) Percent survival Status 20 CR1 15 23.4 CR2 12 18.8 CR у3 3 4.7 PR1 15 23.4 PR у2 3 4.7 25 50 75 100 125 150 Sensitive relapse 8 12.5 Months Primary refractory 8 12.5 Radiotherapy conditioning No 52 81.2 b 100 Yes 11 17.2 Unknown 1 1.6 Source of stem cells Bone marrow 48 75.0 80 Peripheral blood 14 21.9 Both 2 3.1 60

Table 3 Combination of pre-transplant and post-transplant status 40

Status at transplant Percent survival 1st Other PR Sensitive Refractory Total 20 CR CR relapse

(a) Response at 90 days CR 15 15 8 4 42 25 50 75 100 125 150 PR 5 3 2 10 No response 2 4 6 Months Progression 3 2 5 Toxic death 1 1 Figure 1 Overall survival (a) and progression-free survival (b) for all Total 15 15 18 8 8 64 patients. (b) Most recent status CR 14 9 12 1 36 PR 1 1 status at transplant and current status. All the patients Progression 4 4 2 10 = Toxic death 1 1 (n 30) transplanted in CR maintained it at 90 days. Com- Prog and death 1 6 2 2 5 16 plete remission was achieved for 44% of the patients trans- Total 15 15 18 8 8 64 planted in PR and for 50% of the patients transplanted in sensitive relapse. None of those with refractory disease obtained a CR (Table 3a). With respect to the most recent with a CR rate of 50% (details are shown in Table 1). The status, one patient transplanted in first CR (7%), six (40%) median interval time between diagnosis and transplant was of those transplanted in у2nd CR, six (33%) of those trans- 9.6 months (range 3.6–94.6). Status at transplant is detailed planted in PR and seven (87%) of those transplanted in in Table 2: 30 of 64 patients (47%) were transplanted in sensitive relapse progressed (Table 3b). CR, but only 15 in first CR (23% of the total number). Thirty of the 64 cases (47%) responsive to first-line therapy Outcome (CR + PR) were transplanted as consolidation shortly after the end of CHT. Median follow-up of the patients in these series is 43.3 The majority of the patients were conditioned with CHT months. OS of the whole population (n = 64) at 2 years, 5 alone (81%) and were reinfused with bone marrow stem years and 10 years was 76.3%, 70.3% and 70.3% respect- cells (75%). ively (Figure 1). PFS for all patients was 62.4%, 56.4% and 47.0% at 2, 5 and 10 years, respectively, with the median at 83.6 months (Figure 1). Response to treatment Eighteen patients aged less than 20 years were trans- Table 3 illustrates the relationship between status at trans- planted. OS was significantly better (P = 0.032) than that plant and response at 90 days post-transplant and between of the adult population (in fact all patients but one under ASCT for T and null cell CD30-positive ALCL R Fanin et al 440 100 Table 4 Multivariate analyses n = 18, age р20 n = 58 Coefficient Risk 95% CI 80

(a) Survival = 60 Symptoms at diagnosis (1 no; 2.25 9.52 2.41–37.7 n = 46, age >20 2 = yes) Age at transplant (continuous) 0.07 1.08 1.03–1.13 = = 40 No. extranodal sites (0 0; 1 1; 1.05 2.86 1.52–5.41 2 = Ͼ2)

Percent survival Status at transplant 1.12 3.08 1.52–6.26 = = + 20 (1 1st CR, 2 2nd CR PR, 3 = Refract + Prog)

(b) Progression-free survival Stem cells origin (1 = BM; 2 = PB) 1.96 7.13 1.86–27.3 25 50 75 100 125 150 Symptoms at diagnosis (1 = no; 0.92 2.52 1.00–6.36 Months 2 = yes) No. extranodal sites (0 = 0; 1 = 1; 0.41 1.50 −0.03–2.36a р Ͼ Figure 2 Overall survival according to age ( 20 vs 20 years); 2 = Ͼ2) Ͻ P 0.05. Status at transplant 1.16 3.20 0.64–5.42 (1 = 1st CR, 2 = 2nd CR + PR, = + 20 years are still alive) (Figure 2), but there was no differ- 3 Refract Prog) ence in PFS (P Ͼ 0.4). aSignificant in a 1-tail test. Figure 3 shows the outcome (OS) in relation to status at transplant: dividing the patients into four groups (1st CR , 1st PR, other responders and refractory/relapsed) a statisti- NHL to validate the clinical significance of the different cally significant difference emerges (P = 0.006). The out- entities identified by the REAL classification. All types of come at 5 years for patients transplanted in 1st PR and for made up only 12% of the cases, with other responders is similar. only 2.4% of ALCL.11 Harris in 199619 reported a relative Multivariate analysis showed that age at transplant, num- frequency of ALCL in USA and Europe of 5% among lym- ber of extranodal sites at diagnosis, presence of symptoms phomas of the adults and 15% among those of children. at diagnosis and status at transplant were predictive of sur- These data confirm that this is a small group of lymphomas vival (Table 4a). Type of transplant, with PBSC doing and could partially explain the paucity of reports on ALCL worse than BM, symptoms at diagnosis, number of extrano- and ASCT found in the literature. Moreover the majority dal sites and status at transplant were predictive for PFS of the papers published up to now combined the T and null (Table 4b). The AAIPI was of no prognostic value in the cell and the B cell phenotypes that are currently considered prediction of survival or PFS in this group either in multi- as different clinical entities.10 variate analysis or univariate analysis. In this report we concentrate on the T and null cell phenotype. Although no central histology review was per- formed, we have no reason to believe that the histologies Discussion were in error: all centres involved characterise and treat lymphomas regularly, and the accuracy of diagnosis using In 1997, the non-Hodgkin’s Lymphoma Classification Pro- histology and immunophenotyping for this group of ject published a retrospective analysis on 1403 cases of lymphomas is considered to be of 85%.11 It is considered that ALCL does well with conventional 100 therapy and ASCT is not regularly offered to patients with this disease. A review of the literature, however, seems to n = 15, CR1 indicate that the outcome of patients who responded par- 80 tially or were resistant to first-line therapy and of patients n = 15, PR1 n = 18, RES relapsing after the attainment of a CR is not good. In the 60 papers published in the last 5 years1–7 the percent of partial responders is always between 10 and 20%. In the paper by n = 16, R/R Filippa et al,5 out of six partially responding patients, five 40 died and one obtained a CR after ASCT. Longo et al3

Percent survival emphasised that the only significant prognostic factor was 20 the response to initial therapy with an OS at 3 years of patients in CR of 89% vs only 38% of PR/refractory. Penny8 reported three deaths out of six PR. In all these 25 50 75 100 125 150 cases, the longest observation period from diagnosis was Months less than 2 years. The conclusion is that patients who only Figure 3 Overall survival according to status at ASCT: first CR (CR1), respond partially to first-line chemotherapy do not do well. first PR (PR1), other responder (RES), refractory/relapsed (R/R); The prognosis of relapsed patients is also not good: P Ͻ 0.01. Penny et al8 reported four deaths out of seven relapses, ASCT for T and null cell CD30-positive ALCL R Fanin et al 441 of which two had obtained a first CR with conventional In our series 18 patients were younger than 20 years, chemotherapy; Longo et al3 two out of three; Filippa,5 who confirming that T and null cell ALCL are more frequent presented a detailed history of each patient, documented 15 among paediatric and young patients. The survival of this relapses out of 36 patients with an initial CR to chemo- group of patients is significantly superior, while PFS curves therapy. Of these relapses, five (33%) were treated with are identical, suggesting that age does not change the sensi- conventional chemotherapy and were alive at the time of tivity but probably affects the endurance to the treatment. publication although not all were free from disease; seven Age did not influence the outcome of patients who showed (47%) died, and three (20%) were treated with ASCT and a CR 90 days post-ASCT. Younger patients tended to were alive without disease at the time of publication belong to the high risk category as calculated through the (longest observation period 80 months). These data indicate AAIPI despite doing better, further demonstrating the lack not only the poor prognosis of relapsed patients but the of applicability of this index for this group of lymphomas. fairly high incidence of relapses in patients who achieve a Multivariate analysis showed that, aside from status at CR with conventional chemotherapy. transplant, other variables also have a predictive value for With respect to the series of patients in our study, 30 OS and PFS: age with survival, symptoms and number of (46.1%) of them received the procedure in CR with only extranodal sites with both survival and PFS. The better PFS 15 (23%) in first CR. Forty-nine out of 64 patients (76%) of those transplanted with BM rather than with PBSC stem were responding to CHT at the time of transplant. This cells is not easy to interpret, and would require further stud- does not reflect the population that is normally treated with ies with a larger population. conventional chemotherapy stressing that the Registry In conclusion these data seem to indicate that ASCT population is biased towards a higher frequency of worse should be be considered as a possible therapy for ALCL in prognosis patients. complete remission subsequent to first and patients ASCT demonstrated a capacity to improve the quality of responding partially to first-line chemotherapy. If given as the response with a conversion rate of 44% from PR to consolidation, it may also prevent the high level of relapse CR. Moreover the CR rate obtained with transplant was seen until now in patients in first complete remission. maintained. This means that patients with PR after first-line Nevertheless definitive conclusions about the superiority or treatment did better than those reported in the literature. otherwise of high-dose therapy in first-line treatment can ASCT may also have consolidated the response in more only be made through prospective randomised trials. than 50% of the patients who were transplanted in CR sub- sequent to first. OS for patients in CR at 90 days post- ASCT was 91.8% at 2 years; the 10 years projected PFS was 70.7%. The curves tend to a plateau after 24 months Acknowledgements (only one event after this time) confirming that no deaths or relapses are expected after that time-point for these We acknowledge the contribution of the following physicians and patients, as already published by others.20 Transplant does centres that included patients in this study: Hermann F, Abt Innere not apparently change the quality of response in the group Medizin III, Universita¨t Ulm, Germany; Prentice HG, Department of refractory/resistant and relapsed patients and 13 of the of Hematology, Royal Free Hospital, London, UK; Kanz L, Med Univ Klinik, University Hospital, Tu¨bingen, Germany; Goldstone 16 reported deaths happened in this group. AH, Department of Haematology, University College London In support of the possible role of ASCT in salvaging Hospital, UK; Mandelli F, Univ degli Studi La Sapienza, Inst of chemosensitive patients, it has recently been reported that Hematology, Rome, Italy; Ferrant A, Dept of Haematology, Cli- a program of sequential intensive treatment that includes niques Universitaires St Luc, Brussels, Belgium; Verdonck L, CHT Ϯ RT and ASCT as first-line therapy in responding Dept of Hematology, University Hospital, Utrecht, Netherlands; patients (CR + PR) induces a CR rate close to 100% and Iriondo A, Hospital Universitario, ‘Marque´s de Valdecilla’, San- prevents early relapses, projecting all these patients as long- tander, Spain; Dini G, Institute G Gaslini, Genova, Italy; Proctor term survivors.13,14 SJ, Dept of Hematology, Royal Victoria Infirmary, Newcastle- The series of ALCL treated with ASCT reported in our upon-Tyne, UK; Zanesco L, Centro Leucemie Infantili, Padova, study has the longest period of follow-up ever published, Italy; Leone G, Istituto Semeiotica Medica, Ematologia, Univer- sita Cattolica S Cuore, Rome, Italy; Hartmann O, Pediatric BMT with a median observation time from transplant of 3.6 years Unit, Institut Gustave Roussy, Villejuif, France; Ortega J, Hospital (range 3 months–10 years). The actuarial overall survival M Infantil Vall d’Hebron, Barcelona, Spain; Janvier M, Centre of 70.3% (95% CI: 56.1–81.5) at 10 years, given the worse Rene Huguenin, St Cloud, France; Lemonnier MP, Service des stage of disease presented by our patients, seems to be in Malades Sanguines et Tumoral, Groupe Hospitalier Paul-Brousse, excess of what is reported in the literature for patients Villejuif, France; Juli A, Servei d’Hematologia, RG Vall Hebron, treated with conventional chemotherapy (projected follow- Barcelona, Spain; Caillot D, Service Hematologie Adultes, Hopi- up of 5 years).6 tal d’Enfants, Dijon, France; Leverger G, Unite´ d’Hematologie et The outcome of the group under study in the current d’Oncologie Pedia, Hopital Trousseau, Paris, France; Michon J, report could not be predicted from the AAIPI at diagnosis. Institut Curie, Pediatric Oncology UnitWest, Paris, France; Bacca- This lack of applicability of the AAIPI for the T and null rani M, Department of Bone Marrow Transplantation, Udine Uni- versity Hospital, Italy; Koza V, Dept of Hematology/Oncology, cell ALCL is also described in the clinical evaluation that 11 Charles University Hospital, Pilsen, Czech Rep; Caballero, D, S the International Lymphoma Study Group did of NHL Hematologia, Hospital Clinico, Salamanca, Spain; Morris TCM, and is not surprising. In addition, it is not clear that an Belfast City Hospital, Dept of Hematology, UK; Margueritte G, index developed from lymphoma patients undergoing con- Hopital Saint Charles, Unite d’Hemato-Carnologie Pediatrie I, ventional treatment needs be relevant in the ASCT setting. Montpellier, France. ASCT for T and null cell CD30-positive ALCL R Fanin et al 442 References 11 The Non-Hodgkin’s Lymphoma Classification Project. A clinical evaluation of the International Lymphoma Study 1 Greer JP, Kinney MC, Collins RD et al. Clinical features of Group Classification of non-Hodgkin’s lymphoma. Blood 31 patients with Ki-1 anaplastic large-cell lymphoma. J Clin 1997; 89: 3909–3918. Oncol 1991; 9: 539–547. 12 Engelhard M, Brittinger G, Huhn D et al. Subclassification of 2 Shulman LN, Frisard B, Antin JH et al. Primary Ki-1 anaplas- diffuse large B-cell lymphomas according to the Kiel classi- tic large-cell lymphoma in adults: clinical characteristics and fication: distinction of centroblastic and immunoblastic lym- therapeutic outcome. J Clin Oncol 1993; 11: 937–942. phomas is a significant prognostic risk factor. Blood 1997; 89: 3 Longo G, Federico M, Pieresca C et al. Anaplastic large cell 2291–2297. lymphoma (CD30+/Ki-1+). Analysis of 35 cases followed at 13 Fanin R, Silvestri F, Geromin A et al. Primary systemic CD30 GISL centres. Eur J Cancer 1995; 31: 1763–1767. (Ki-1)-positive anaplastic large cell lymphoma of the adult: 4 Clavio M, Rossi E, Truini M et al. Anaplastic large cell lym- sequential intensive treatment with the F-MACHOP regimen phoma: a clinicopathologic study of 53 patients. Leuk Lym- (Ϯ radiotherapy) and autologous bone marrow transplantaion. phoma 1996; 22: 319–327. Blood 1996; 87: 1243–1248. 5 Filippa DA, Ladanyi M, Wollner N et al. CD30 (Ki-1)-posi- 14 Fanin R, Silvestri F, Geromin A et al. Sequential intensive tive malignant lymphomas: clinical, immunophenotypic, histo- treatment with the F-MACHOP regimen (Ϯ radiotherapy) and logic, and genetic characteristics and differences with Hodg- autologous stem cell transplantation for primary systemic kin’s disease. Blood 1996; 87: 2905–2917. CD30 (Ki-1)-positive anaplastic large cell lymphoma of the 6 Zinzani PL, Bendandi M, Martelli M et al. Anaplastic large- adults. Leuk Lymphoma 1997; 24: 369–377. cell lymphoma: clinical and prognostic evaluation of 90 adult 15 Oken MM, Creech RH, Tormey DC. Toxicity and response patients. J Clin Oncol 1996; 14: 955–962. criteria of the Eastern Cooperative Oncology Group. Am J 7 Tilly H, Gaulard P, Lepage E et al for the Groupe d’Etudes Clin Oncol 1982; 5: 649–655. des Lymphomes de l’Adulte. Primary anaplastic large cell 16 Kaplan JO, Meier P. Nonparametric estimation from incom- lymphoma in adults: clinical presentation, immunophenotype plete observation. J Am Stat Assoc 1958; 53: 457–481. and outcome. Blood 1997; 9: 3727–3734. 17 Cox DR. Regressions models and life tables. JR Stat Soc 1983; 8 Penny RJ, Blaustein JC, Longtine JA et al. Ki-Positive large 34: 187–220. cell lymphomas, a heterogeneous group of neoplasms. Cancer 18 The International Non-Hodgkin’s Lymphoma Prognostic Fac- 1991; 68: 362–373. tors Project. A predictive model for aggressive non-Hodgkin’s 9 Pileri S, Bocchia M, Baroni CD et al. Anaplastic large cell lymphoma. New Engl J Med 1993; 14: 987–994. lymphoma (CD30+/Ki-1+): results of a prospective clinico- 19 Harris NL. Principles of the revised European–American lym- pathological study of 69 cases. Br J Haematol 1994; 86: phoma classification (from the International Lymphoma Study 513–523. Group). Ann Oncol 1997; 8: 11–16. 10 Harris NL, Jaffe ES, Stein H et al. A revised European–Amer- 20 Pettengell R, Radfod JA, Morenstern GR. Survival benefit ican classification of lymphoid neoplasms: a proposal from from high-dose therapy with autologous blood progenitor-cell the International Lymphoma Study Group. Blood 1994; 84: transplantation in poor-prognosis non-Hodgkin’s lymphoma. J 1361–1392. Clin Oncol 1996; 14: 586–592.