Age-Adapted Induction Treatment of Acute Lymphoblastic Leukemia in the Elderly and Assessment of Maintenance with Interferon Combined with Chemotherapy

Home , Daunorubicin

Age-adapted induction treatment of acute lymphoblastic leukemia in the elderly and assessment of maintenance with interferon combined with chemotherapy. A multicentric prospective study in forty patients A Delannoy1, C Sebban2, P Cony-Makhoul3, B Cazin4, C Cordonnier5, R Bouabdallah6, JY Cahn7, F Dreyfus8, A Sadoun9, JP Vernant5, C Gay2, A Broustet3, JL Michaux1 and D Fie`re2 for the French Group for Treatment of Adult Acute Lymphoblastic Leukemia

1Groupe d’He´matologie de l’Universite´ Catholique de Louvain, Belgium; and Departments of Hematology of 2Hoˆ pital Edouard Herriot, Lyon; 3Hoˆpital Haut Le´veˆque, Bordeaux; 4CHU Lille; 5Hoˆ pital Henri Mondor, Cre´teil; 6Hoˆ pital Paoli-Calmette, Marseille; 7Hoˆpital Jean Minjoz, Besanc¸on; 8Hoˆ pital Cochin, Paris; 9CHU Poitiers, France

Acute lymphoblastic leukemia (ALL) in the elderly is charac- by anecdotal reports suggesting that IFN may prove effective terized by its poor prognosis. Forty patients with ALL, aged 55 in some refractory or relapsing ALL.10–14 Moreover, in patients years or older, and with good performance status (ECOG Ͻ3) receiving IFN after bone marrow transplantation, the risk for were prospectively treated according to an age-adapted regi- 15 men: induction therapy was derived from the LALA87 protocol subsequent relapse of leukemia is reduced and, finally, a while the feasibility of treatment with interferon combined with hemizygous or homozygous deletion of the IFN gene has been chemotherapy was assessed during maintenance. Compared reported in 30% of ALL patients, which led to the hypothesis with younger adults treated according to the LALA87 protocol, that a functional or quantitative defect of endogenous IFN may elderly patients did not present with more adverse prognostic be involved in the pathogenesis of ALL.16 features, except for a lower incidence of T cell ALL (9 vs 31%, P = 0.005). There were even less patients with a high leukocyte We present an analysis of this treatment protocol in 40 con- count (15 vs 38%, P = 0.003), a characteristic associated with secutive ALL patients aged 55 years or older. The limit of 55 adverse prognosis while the incidence of Philadelphia-positive years of age was elected in accordance with our current treat- (Ph-positive) ALL was not significantly increased compared to ment policy in adult patients who are ineligible for high-dose younger adults (31 vs 20%, P = 0.2). After completion of induc- therapy and stem cell transplantation beyond 55 years. In tion therapy, with or without salvage treatment, 85% (CI: 70– addition, the clinical and biological characteristics of elderly 94%) obtained a complete response (CR) while treatment- related mortality during induction was 7.5% (CI: 2–20%). patients with ALL are compared with those of younger adults Median overall survival and disease-free survival were 14.3 treated according to the LALA87 protocol between 1987 months and 14 months, respectively, which, although inferior and 1991.9 to results achieved in younger adults, compares favorably with available data in the elderly. Treatment with IFN proved feasible in most patients but had to be discontinued in eight patients Patients and methods because of toxicity. Age-adapted treatment improves the prognosis of ALL in the elderly even if, in most cases, a cure cannot be achieved. Patients Keywords: acute lymphoblastic leukemia; interferon; elderly; prognosis; chemotherapy From March 1992 to February 1995, 40 patients from 11 French and Belgian Hematology departments were included in this study. All of them had previously untreated ALL and Introduction were aged 55 years or older. Patients with a poor performance status (ECOG 3–4) or with another concomitant malignant Only few reports specifically devoted to acute lymphoblastic condition or with a poor cardiac function preventing the use leukemia (ALL) in the elderly are available,1–7 probably of anthracyclines as assessed clinically or by ventricular ejec- because the condition is uncommon. Although heterogeneous tion fraction measurement were not eligible. with regard to population characteristics, these studies consist- ently report poor median survival ranging from 31 to 10 months5 with a high death rate of up to 50%, during induction Immunophenotypic classification therapy.8 Moreover, shorter median remissions (12 months) are obtained in older adults than in younger patients (24 Classification of ALL according to immunophenotype was as + + months) treated according to similar protocols.5 follows: B-ALL (CD19 or CD20 , surface immunoglobulin + + − In 1992, the French Group for Treatment of Adult ALL +ve), Pre B-ALL (CD19 or CD20 , CD10 ), common ALL + + + + + initiated a study aimed at improving the prognosis of ALL in (CD19 or CD20 , CD10 ), T-ALL (CD2 or CD3 ). older patients. Induction therapy was derived from the young adult protocol LALA879 with chemotherapy administered at lower dose and tailored according to the response assessed Treatment on day 15. In addition, the feasibility of maintenance with interferon (IFN) combined with chemotherapy was assessed. The treatment scheme is presented in Figure 1. Compared The introduction of IFN in the treatment of ALL is supported with the LALA87 protocol used by us in younger adults, its main characteristics are: lower dose of chemotherapy, prednisone administered on alternate day, number of courses of Correspondence: A Delannoy, Cliniques St Luc, Hematology Depart- anthracyclines tailored according to response as assessed by ment, Av. Hippocrate, 1200 Brussels, Belgium day 15 bone marrow study, and optional use of growth factors Received 10 April 1997; accepted 23 April 1997 during induction therapy. Consolidation therapy included all Acute lymphoblastic leukemia in the elderly A Delannoy et al 1430 Comparison with younger adult patients

The clinical and biological characteristics of this cohort were compared to those from 15- to 55-year-old patients treated according to the LALA87 protocol.9

Statistics

Data were analyzed in December 1995. Proportions were compared using Fisher’s exact probability test. Survival was calculated from the day of diagnosis using the method of Kaplan and Meier until death or date last known alive.18 Dis- ease-free survival (DFS) was calculated from the date of complete remission until relapse, death or date last known alive. The log-rank test was used in order to compare survival of Figure 1 Treatment scheme. Induction: vincristine (2 mg weekly subsets of patients. Ninety-five percent confidence intervals for 4 weeks), cyclophosphamide (400 mg/m2 weekly for 4 weeks), are abridged to CI. prednisone 60 mg/m2 on alternate days from day 1 to day 22 and daunorubicin 30 mg/m2 weekly for 4 weeks. In patients with less than 20% blast cells in the bone marrow specimen examined on day 15, the third and fourth injections of daunorubicin were omitted. A bone Results marrow study was performed by day 28. For patients obtaining a complete response (CR), consolidation treatment was scheduled to start at Patient characteristics day 35 while in patients not achieving CR by day 28, salvage with consolidation treatment was started immediately. Consolidation ther- The main patient characteristics are reported in Table 1 and 2 2 apy: daunorubicin 40 mg/m on day 1, cytosine arabinoside 60 mg/m are compared to similar data in younger adults treated accord- intramuscularly (i.m.) or intravenously (i.v.) from day 1 to day 5, L- asparaginase 500 U/kg i.m. or i.v. daily from day 6 to day 10. A ing to the LALA87 protocol. The median age was 66.8 years second course of daunorubicin, cytosine arabinoside and asparagin- (range: 55–83). Compared with younger adults, our patients ase was administered to patients achieving CR after salvage therapy. presented less often a high leukocyte count (median WBC Treatment of occult central nervous system (CNS Tx) involvement: count: 6.8 × 109/l, range: 0.6–160) and T cell ALL. Other two intrathecal injections of methotrexate (10 mg/m2) during induc- characteristics, including organomegaly, thrombocytopenia, tion therapy, two injections during the consolidation phase and one CNS involvement and Ph-positive ALL, were similar to those when maintenance was started. CNS radiotherapy was performed only in patients with overt CNS leukemic involvement. Maintenance: observed in younger adults. Immunological subtypes (B-ALL, methotrexate, 6 mg/m2 i.m. weekly, 6-mercaptopurine, 20 mg/m2 pre-B ALL, Common ALL, T-ALL) were as follows: one, 15, 14 daily, and IFN (interferon ␣-2b, Intron A, Schering Plough, Kenilworth, and three patients, respectively. Immunophenotyping was not NJ, USA) 3 MU/m2 subcutaneously, thrice weekly, during 2 years. available in four patients while three patients could not be Maintenance treatment was started immediately after the consoli- assigned to a clear-cut immunological subtype. dation phase provided the neutrophil count was above 1.5 × 109/l and the platelet count above 125 × 109/l. Doses were adapted according to hematologic toxicity as follows: 6-mercaptopurine was discontinued if the platelet count was lower than 125 × 109/l or if the neutro- Response to treatment phil count was lower than 2.5 × 109/l. When the platelet count was below 75 × 109/l or when the neutrophil count was lower than At completion of induction therapy, 31 complete responses 1.5 × 109/l, all three drugs used during maintenance were discon- were achieved (77.5%, CI: 62–89%). After a first consoli- tinued and, when blood counts improved, were resumed at half the dation course, three additional patients obtained a CR. Thus, initial dose and subsequently at full dose if blood counts permitted. 34/40 patients (85%, CI: 70–94%) obtained a CR. A bone marrow examination was performed in 36 patients at day 15: in 21 patients, the blast cell count was less than 20% while in 15 patients more than 20% blast cells could be demonstrated. the drugs used in the LALA87 protocol but at lower doses. Two additional injections of daunorubicin were administered Cranial irradiation was performed only in patients with central in nine of the latter 15 patients, as stated in the treatment pro- nervous system (CNS) leukemic involvement. IFN was intro- tocol, while in six patients, the two courses of daunorubicin duced during maintenance, in addition to methotrexate and were omitted because of their poor condition. 6-mercaptopurine. Out of the 34 complete responders, 32 patients received the scheduled consolidation therapy (one patient with a normal bone marrow died before consolidation was started and another one was felt unfit to proceed with consolidation Assessment of response treatment). Five patients relapsed during the consolidation phase. Thus, 27 patients were alive and in CR after consolidation Complete response definition was derived from the CALGB’s therapy. Twenty-five proceeded with maintenance treatment criteria: bone marrow with Ͻ5% blast cells, neutrophil count as stated in the protocol: one patient refused treatment with Ͼ1 × 109/l, platelet count Ͼ100 × 109/l for at least 30 days.17 IFN while IFN was not administered in another patient Relapse was defined as the appearance of leukemic cells in because of psychiatric disorder. Relapse occurred in 18 the bone marrow (Ͼ5%) and/or reappearance of clinical and patients during maintenance therapy 81–761 days after IFN hematological evidence of disease. was started (median 237 days). Acute lymphoblastic leukemia in the elderly A Delannoy et al 1431 Table 1 Patient characteristics at diagnosis and comparison with younger patients treated according to the LALA87 protocol

This cohort Younger adults (9)a P n 40 525 Median age and range (years) 66.8 (55–86) 30 (15–55) Male/female (ratio) 20/20 (1) 328/197 (1.6) 0.13 Mediastinal mass (%) 4/40 (10) 78/507 (15) 0.49 Hepatomegaly (%) 10/40 (25) 152/520 (29) 0.7 Splenomegaly (%) 13/40 (32.5) 263/520 (50) 0.03 CNS involvement (%) 3/40 (7.5) 35/506 (7) 0.75 Leukocyte count Ͼ30 × 109/l (%) 6/40 (15) 201/525 (38) 0.003 Platelet count Ͻ100 × 109/l (%) 30/40 (75) 360/525 (68) 0.47 T-ALL (%) 3/33b (9) 148/473 (31) 0.005 Ph-positive ALL (%) 8/26c (31) 48/241 (20) 0.2 aOnly patients aged Ͻ55 years are taken into account in this table. bThe seven patients with unknown or unclassiﬁable phenotype are not taken into account. cPhiladelphia chromosome could be assessed by cytogenetic or/and molecular analysis in 26 patients.

Toxicity from induction and consolidation treatment reductions were mainly due to regimen-induced hematologic toxicity. Three patients died early from treatment-related infectious complications: the first during induction treatment, another one died during the consolidation phase after failing induc- Overall survival and disease-free survival tion, while an additional patient with a normal bone marrow died shortly after completing induction therapy. During induc- The median survival of the whole group was 434 days and Ͻ × 9 tion, neutropenia (neutrophil count 0.5 10 /l) lasted from the median DFS was 427 days (Figures 3 and 4). In this small 0 to 57 days (median: 20 days) while a thrombocytopenia cohort, no survival difference could be demonstrated accord- Ͻ × 9 50 10 /l was recorded during 0 to 60 days (median 15 ing to sex, age (Ͼ70 vs Ͻ70 years), anemia, leukocyte count, days). Growth factors (G-CSF) were administered in only two platelet count, normal vs abnormal chromosome analysis, patients. Aplasia from consolidation therapy was mild with presence or absence of organomegaly. In addition, survival Ͻ × 9 both neutropenia 0.5 10 /l and thrombocytopenia did not correlate with day 15 clearance of bone marrow Ͻ × 9 50 10 /l lasting 0–17 days (median: 0). malignant cells: median survival of 21 patients with Ͻ20% blast cells in day 15 bone marrow specimen was 425 vs 434 days in 15 patients with Ͼ20% blast cells (P = 0.6). Also, the Feasibility of maintenance with IFN CR rate was not affected by day 15 blast count with 19/21 and 11/15 CR being obtained in patients with Ͻ20% and Of the 25 patients who proceeded with maintenance therapy, Ͼ20% bone marrow blast cells by day 15, respectively. Mov- eight suffered complications leading to discontinuation of IFN: ing the cutoff value of day 15 blast cell count from 20 to 5% severe anorexia, unacceptable hemototoxicity and mood dis- as recommended by Sebban et al19 also failed to distinguish orders with anorexia developed in four, one and two patients, patients with a favorable outcome (median survival 364 vs respectively. In addition, IFN was withheld in one patient who 543 days in 16 patients with Ͻ5% blast cells and in 20 developed hemiplegia, possibly unrelated to therapy, 10 days patients with Ͼ5% blast cells, respectively, P = 0.3), again after starting IFN. Effectively provided doses, expressed as a with no impact on CR rates (14 CR in 16 patients with Ͻ5% percentage of the theoretical dose, are reported in Figure 2. bone marrow blast cells vs 16 CR in 20 patients with a blast On the average, three quarters of the theoretical doses of both cell count Ͼ5%). The median survival of the nine patients IFN and chemotherapy could be administered. Dose who were administered two additional courses of daunorubicin because of day 15 bone marrow infiltration with Ͼ20% blast cells was 262 days, vs 425 days in the 21 patients with Ͻ20% blast cells (P = 0.1). The median survival of the six patients who were felt unfit for receiving two additional injections of daunorubicin was 740 days (P = 0.1 compared with the nine patients given additional daunorubicin). Relapse occurred in 23 patients, including 16 bone marrow relapses, five CNS relapses and two CNS associated with bone marrow relapses. Two-year overall survival and DFS were 20 and 16%, respectively.

Discussion Figure 2 Mean doses of interferon (IFN), methotrexate (MTX) and 6-mercaptopurine (6-MP) and their standard deviation assessed quar- Although more encouraging than previous studies, this report terly as a percentage of the theoretical dose during the 8-quarter main- confirms the poor prognosis of elderly patients with ALL. As tenance. a comparison, the median survival of younger adult patients Acute lymphoblastic leukemia in the elderly A Delannoy et al 1432 in elderly people with ALL and to assess the feasibility of therapy with IFN during maintenance. With regard to outcome, Table 2 compares our data with those previously reported in patients accrued in treatment protocols. Thus, population- based studies were not taken into account since, in such studies, all patients, including those not given chemotherapy, are included, which results in poor overall survival rates compared with survivals obtained in the setting of treatment protocols. In addition, patients aged 60 years and older are shown separately, as only patients aged 60 years or older were included in most other studies. Although comparisons should be interpreted with great caution, data from Table 2 suggest that more complete responses could be achieved in our patients with less toxic deaths which translated into longer overall survival. Figure 3 Overall survival. Median survival was 434 days. Two- Treatment protocols designed for young adults when used = year survival 20%. unmodified in older patients result in prohibitive mortality.8 On the other hand, when induction treatment is oversim- treated according to the LALA87 protocol was 18 months vs plified, (eg prednisone-vincristine), toxicity is mild but efficacy 4,6 14 months in the present study9 while 3-year overall survival is also very modest. In this study, we elected to apply an was 36% compared with 10% in older adults. The reasons for intermediate treatment policy: elderly patients were adminis- a better survival in younger adults with ALL are not clearly tered all the drugs used by our group in younger adults but understood. A high leukocyte count, CNS involvement, organ- doses were lowered. In addition, treatment was tailored omegaly, which are poor prognosis characteristics, were not according to response as assessed by bone marow study at encountered more frequently in elderly patients. On the con- mid-treatment which may have contributed to improving trary, the leukocyte count was lower in our patients compared overall results. Recent data suggest that in adults with ALL, to younger adults. T cell ALL comprises 20–25% of adult ALL day 15 marrow is predictive both for response and for sur- 19 and has been associated with a favourable prognosis.8 Rela- vival. Using the same cutoff of 5% blast cells in day 15 bone tively few patients suffered T-ALL in his cohort (9%), which marrow specimens we could not confirm the prognostic value might have adversely affected their prognosis. A relatively low of day 15 bone marrow findings in this relatively small cohort. frequency of T-ALL in older patients has been reported by We suggest that the comparatively encouraging outcome others: 0/15,1 2/41,2 0/9,3 2/26.4 However, the poor prognosis observed in our patients results from treatment-related charac- of ALL in older persons cannot be accounted for only by a teristics, although, outside the setting of a randomized trial, low incidence of T-ALL as the prognosis of ALL in the elderly alternative explanations, including study population charac- is poor, even when compared to the prognosis of B-ALL in teristics, cannot be ruled out. Of note, our patients did not younger patients. Philadelphia chromosome has been consist- differ from those treated elsewhere with respect to age and to ently associated with a poor outcome in ALL. The incidence the proportion of Ph-positive ALL, two main determinants of of Ph-positive ALL increases with age but seems to stabilize outcome in ALL, while, conversely, excluding from our study beyond the age of 60.20 This was also our experience: patients with a poor cardiac function or performance status, although more elderly patients presented with Ph-positive leu- and with another simultaneous malignancy may have resulted kemia (31%) compared to younger adults treated according in selecting a subset of patients with favorable prognostic fea- to the LALA87 protocol (20%), the difference was not signifi- tures compared to studies presented in Table 2. cant and could not account on its own for the poor prognosis Hematotoxicity from consolidation therapy in this study was of ALL in the elderly. very mild. Thus, although the therapeutic potential of consoli- The main purposes of our study were to improve outcome dation treatment remains controversial, higher doses of chemotherapy could be safely used in future protocols. Sur- vival could be further increased by intensifying CNS prophy- laxis as an unduly high number of CNS relapses (7/23), reasonably ascribable to our reduced CNS therapy, occurred in this cohort. With regard to the feasibility of using IFN and chemotherapy during maintenance therapy, most elderly patients could tolerate IFN provided the dose was reduced to three quarters of the planned dose. This study was not designed to assess the efficacy of IFN during maintenance. However, similar to previously reported cohorts, few patients if any will be cured which suggests that IFN is unlikely to improve survival dramatically. Finally, we cannot exclude that reducing the doses of methotrexate and 6-mercaptopurine may have offset the anti-leukemic effect of IFN, if any. In conclusion, although prognosis in older patients with ALL remains poor, age-adapted treatment results in CR rates similar to those achieved in younger adults. If higher doses of Figure 4 Disease-free survival in 34 complete responders. Median chemotherapy can be safely administered in the elderly when DFS is 427 days. Two-year DFS = 16%. growth factors are systematically used remains to be demon- Acute lymphoblastic leukemia in the elderly A Delannoy et al 1433 Table 2 Comparison between our data and previously reported studies. Only studies reporting on patients included in treatment protocols are presented here

Author/Ref. n Median age Ph-positive Treatment CR (%) Death rate Median Median (years) ALL (%) during survival disease-free induction (months) survival (%) (months) or median CR duration (*)

Hussein8 40 NS (all Ͼ50 17 V-P-A 35 50 1 8.3 years old) Taylor2 19 68 Misc. 37 15 3 NS Spa¨th-Schwalbe4 28 65 Misc. 43 37 5 NS Kantarjian5 52 NS (all Ͼ60 VAD 65 12 10 12(*) years old) Ferrari6 49 68 V-P-D-Asp or V-P 59 23 9 7(*) Bassan7 22 64 43 Ida-V-P-Asp 59 18 5.3 NS This study 40 66.8 31 V-P-D-Cy 85 7.5 14.2 14 This study (patients aged 34 67.7 27 82 6 14.2 10 60 or older)

NS, not stated; V, vincristine; P, prednisone; A, adriamycin; Misc. miscellaneous; VAD, vincristine, adriamycin, dexamethasone, Asp, L-asparaginase; D, daunorubicin; Ida, idarubicin; Cy, cyclophosphamide. strated, although preliminary data seem encouraging.21 More 9 Fie`re D, Lepage E, Sebban C, Boucheix C, Gisselbrecht C, Vernant intensive consolidation and CNS treatments are being investi- JP, Varet B, Broustet A, Cahn JY, Rigal-Huguet F, Witz F, Michaux gated by our group with the hope of improving remission dur- JL, Michallet M, Reiffers J for the French Group on Therapy for Adult Lymphoblastic Leukemia. Adult acute lymphoblastic leuke- ation while the therapeutic value of IFN as a post-induction mia: a multicentric randomized trial testing bone marrow trans- therapy is being prospectively assessed. plantation as postremission therapy. J Clin Oncol 1993; 11: 1990–2001. 10 Einhorm S, O¨ st Å, Grimfors G, Grande´r D, Bjo¨rkholm M. Pro- Acknowledgements longed remission induced by interferon in a patient with refractory T-cell ALL. Lancet 1990; i: 794–795. 11 Ochs J, Brecher ML, Mahoney D, Vega R, Pollock BH, Buchanan We wish to thank Schering Plough France and Belgium for GR, Whitehead VM, Ravindranath Y, Freeman AI. Recombinant their support and for supplying interferon. interferon alfa given before and in combination with standard chemotherapy in children with acute lymphoblastic leukemia in first marrow relapse: a Pediatric Oncology Group pilot study. J References Clin Oncol 1991; 9: 777–782. 12 Makhonova LA, Gavrilova IE, Maiakova SA, Kuznetsov VP, Orlova 1 Delannoy A, Ferrant A, Bosly A, Chatelain C, Doyen C, Martiat TG, Kadagidze ZG. Results of the use of human leukocytic inter- P, Michaux JL, Sokal G. Acute lymphoblastic leukemia in the eld- feron in the treatment program in acute lymphoblastic leukemia erly. Eur J Hematol 1990; 45: 90–93. in children. Gematol Transfuziol 1989; 34: 33–36. 2 Taylor PRA, Reid MM, Bown N, Hamilton PJ, Proctor SJ. Acute 13 Opat B, Drbal J, Vyzula R. High dose alpha-2-recombinant inter- lymphoblastic leukemia in patients aged 60 years and over: a feron in the treatment of malignant lymphoproliferative diseases. population-based study of incidence and outcome. Blood 1992; A pilot study. Neoplasma 1989; 36: 489–494. 80: 1813–1817. 14 Lauer SJ, Ochs J, Pollock BY, Buchanan GR. Recombinant alpha 3 Virgilio JF, Moscinski LC, Ballester OF, Corrado C, Guida C, Bald- 2B interferon treatment for childhood T-lymphoblastic disease in ucci L, Saba H. Acute lymphocytic leukemia in elderly patients. relapse. A Pediatric Oncology Group Phase II study. Cancer 1994; Hematol Oncol 1993; 11: 121–126. 74: 197–202. 4 Spa¨th-Schwalbe E, Heil G, Heimpel H. Acute lymphoblastic leu- 15 Meyers JD, Flournoy N, Sanders JE, McGuffin RW, Newton BA, kemia in patients over 59 years of age. Experience in a single Fisher LD, Lum LG, Appelbaum FR, Doney K, Sullivan KM, Storb center over a 10-year period. Ann Hematol 1994; 69: 291–296. R, Buckner D, Thomas ED. Prophylactic use of human leukocyte 5 Kantarjian HM, O’Brien S, Smith T, Estey EH, Beran M, Preti A, interferon after allogeneic marrow transplantation. Ann Intern Med Pierce S, Keating MJ. Acute lymphocytic leukaemia in the elderly: 1987; 107: 809–816. characteristics and outcome with the vincristine-adriamycin- 16 Diaz MO, Rubin CM, Harden A, Ziemin S, Larson RA, Le Beau dexamethasone (VAD) regimen. Br J Haematol 1994; 88: 94–100. MM, Rowley JD. Deletions of interferon genes in acute lympho- 6 Ferrari A, Annino L, Crescenzi S, Romani C, Mandelli F. Acute blastic leukemia. New Engl J Med 1990; 322: 77–82. lymphoblastic leukemia in the elderly: results of two different 17 Ellison RR, Holland JF, Weil M. Arabinosyl cytosine: a useful agent treatment approaches in 49 patients during a 25-year period. Leu- in the treatment of acute leukemia in adults. Blood 1968; 32: kemia 1995; 9: 1643–1647. 507–523. 7 Bassan R, Di Bona E, Lerede T, Pogliani E, Rossi G, D’Emilio A, 18 Kaplan EL, Meier P. Non parametric estimation from incomplete Buelli M, Rambaldi A, Viero P, Rodeghiero F, Barbui T. Age- observations. J Am Stat Assoc 1958; 53: 457–481. adapted moderate-dose induction and flexible outpatient post- 19 Sebban C, Browman GP, Lepage E, Fie`re D. Prognostic value of remission therapy for elderly patients with acute lymphoblastic early response to chemotherapy assessed by the day 15 bone mar- leukemia. Leuk Lymphoma 1996; 22: 295–301. row aspiration in adult acute lymphoblastic leukemia: a prospec- 8 Hussein KK, Dahlberg S, Head D, Waddell CC, Dabich L, Weick tive analysis of 437 cases and its application for designing induc- JK, Morrison F, Saiki JH, Metz E, Rivkin SE, Grever MB, Boldt B, tion chemotherapy trials. Leukemia Res 1995; 19: 861–868. and the Southwest Oncology Group. Treatment of acute lym- 20 Groupe Français de Cytogeńe´tique He´matologique. Cytogenetic phoblastic leukemia in adults with intensive induction, consoli- abnormalities in adult acute lymphoblastic leukemia: correlations dation, and maintenance chemotherapy. Blood 1989; 73: 57–63. with hematologic findings and outcome. A collaborative study of Acute lymphoblastic leukemia in the elderly A Delannoy et al 1434 the Groupe Français de Cytogeńe´tique He´matologique. Blood (Filgrastim: G-CSF) reduces the time to neutrophil recovery in 1996; 87: 3135–3142. adults with acute lymphoblastic leukemia receiving intensive 21 Larson RA, Linker CA, Dodge RK, George SL, Davey FR, Frankel remission induction chemotherapy: Cancer and Leukemia Group SR, Powell BL, Schiffer CA. Granulocyte-colony stimulating factor B Study 9111. Proc Am Soc Clin Oncol 1994; 13: 305a (Abstr.).