Randomized study of recombinant interleukin-2 after autologous bone marrow transplantation for acute leukemia in first complete remission.

ARTICLE

INTRODUCTION

Despite continuous achievements over the last 20 years in biological understanding and therapeutic approaches, patients suffering from acute leukemias usually have a poor prognosis. Despite the fact that initial disease can usually be controlled by cytotoxic chemotherapy, most patients will relapse and eventually die. Thus, attempts to prevent leukemia recurrence have been the target of ongoing investigations. Much effort has been directed towards intensification of chemotherapy with or without stem cell support as a way to consolidate first remission [1-3]. Despite some significant improvements, especially in improving remission duration, relapse remains, nevertheless, the major cause of death in more than 50% of patients. On the other hand, one promising approach has been the use of allogeneic bone marrow transplantation (BMT), which represents a fascinating alternative. Although final survival is still impaired by transplant-related mortality, it currently represents the best method to control and eventually eradicate residual disease for patients in remission. It has been convincingly demonstrated that control of minimal residual disease is reinforced by allogeneic lymphoid cells contained in the graft which exert the so-called graft versus leukemia effect (GVL). Details of GVL mechanism however have not yet been totally elucidated, although it is recognized that it represents an immunological reaction of the donor immune system.

In the late eighties, following Steven Rosenberg's work in solid tumors sensitive to immunological control, we investigated the role of immunotherapy in the treatment of acute leukemias (AL) with recombinant interleukin-2 (r-IL-2). We reported that r-IL-2 was able to induce complete remission in selected patients with very advanced leukemias [4, 5]. We showed that this treatment was associated with the impairment of some normal cell functions (granulocytes) [6, 7], and with the modification of the adhesion molecule expression on the surface of the blasts [8, 9]. We hypothesized that this latter information suggested a cellular pathway for the mechanism of the antileukemic effect. Other groups also reported similar effects of r-IL-2 on AL [10, 11]: in particular Meloni et al. reported prolonged responses in less proliferative leukemias [12]. Altogether, these data demonstrated that acute leukemias, especially acute myeloblastic leukemias (AML) may be sensitive to the immunological effects exerted by r-IL-2 therapy and, warranted investigation to define the most appropriate setting for use.

Autologous (auto) BMT appears to be an appealing situation in which to assess the efficacy of r-IL-2 in acute leukemias. Auto-BMT combines the setting of minimal residual disease with a specific immune situation where activated cytotoxic T and NK cells are spontaneously generated [13], with both elements possibly favoring successful immunotherapy. In fact, additional immune cell stimulation and activation with r-IL-2 may represent the best method of enhancing the antileukemic effect on minimal residual disease with the hope of reaching the level of cure obtained after allogeneic BMT. We thus first established that r-IL-2 could be used after autologous BMT with acceptable side effects [11, 14, 15]. In these early feasibility studies, we also showed that the r-IL-2 treatment generated a significant immune stimulation [14, 15] as well as clinical events similar to those seen after allogeneic BMT [16, 17]. In order to define whether such a strategy had efficacy in the treatment of AL we designed a multicenter phase III study investigating the role of r-IL-2 in relapse prevention after auto-BMT for AL. Patients with AL were treated following an auto-BMT performed as consolidation of the first complete remission (CR1): we report here the results of this study which includes 130 patients with a minimal follow-up of five years.

PATIENTS AND METHODS

Patients

Patients included were first randomized between two groups receiving (study group) or not receiving (control group) r-IL-2) and then underwent auto-transplantation. Patients with AML or lymphoblastic (ALL) leukemia treated with auto-BMT in CR1 were eligible for this trial. In order to assess the feasibility of this strategy, randomization was performed at the time of transplantation. Thirteen centers included patients in this study. Some patients in this study have previously been reported [18, 19].

The protocol was reviewed and approved by the CCPPRB (Committee of Ethics) of Marseille. Written informed consent was obtained from all patients or their legal guardians. Bone marrow was harvested at the time of first complete remission and was frozen without in vitro purging. It was re-infused after a preparative regimen consisting of cyclophosphamide (CY) (120 mg/kg) and total body irradiation (TBI) with a minimal cumulative dose of 10 Gy.

r-IL-2 was started when hematological recovery was obtained, defined as an absolute granulocyte count >= 0.5 x 10⁹/l and self-sustaining platelets >= 50 x 10⁹/l. Patients were readmitted in order to receive r-IL-2 on a conventional hospital ward. All patients were treated with the same schedule of r-IL-2 as previously described [20], which consisted of a continuous infusion of r-IL-2, but at a dose of 12 x 10⁶ IU/m² for 5 cycles (RU 49637, provided by Dr. Brandely, Roussel-Uclaf, Romainville, France). The first cycle, starting by convention on Day 1, lasted a maximum of 5 days and was followed by 4 cycles of 2-day treatment, each started on day 15, 29, 43 and 57. This schedule allowed a maximum of 13 days of treatment over a period of 60 days, with a theoretical total dose of r-IL-2 of 156 x 10⁶ IU/m². Patients were routinely medicated to alleviate symptoms during r-IL-2 therapy as previously reported [20]. No patients received corticosteroids, but all received prophylactic antibiotics consisting of Pefloxacin (200 mg x 2/day) and penicillin (1 x 10⁶ Units x 2/day), during the whole treatment period. No IV fluid was administered as part of r-IL-2 infusion, but furosemide was given, if needed, to avoid symptomatic fluid retention. Support of hypotension and oliguria consisted of volume replacement with 4% human albumin. Dopamine was then added, if necessary, at low doses (2.5 mug/kg/min) to maintain renal perfusion. If hypotension require vasopressor medication, in addition to albumin infusion, dopamine was increased to 5 to 10 mug/kg/min and r-IL-2 was discontinued. Patients who developed anemia were transfused with irradiated (15 Grays) packed red blood cells (PRBC) to maintain hemoglobin (HB) levels above 10 g/100 ml and, if necessary, irradiated platelet transfusions were given to maintain platelet levels above 20 x 10⁹/l. Vital signs were checked every 4 hours during the r-IL-2 administration. Toxicity was graded according to the World Health Organization (WHO) scale. r-IL-2 was stopped when a grade 3 or several simultaneous grade 2 toxicities occurred, until they had completely resolved. r-IL-2 could then be started again, if necessary with a 50% decrease in dose, when it was considered unlikely that toxicities would recur. Patients were discharged when all toxicities had resolved.

Biological evaluation

All patients had standard blood chemistry and hematological counts, checked on a daily basis from the day before to the day after the end of r-IL-2 infusion. Between the r-IL-2 infusions, checks were done twice a week. Cytogenetic abnormalities were classified into three groups as previously published [20, 21].

Statistical evaluation

All data were computed using SPSS for Windows software, Chicago, IL60611, USA. The Mann and Whitney U test was used to test differences in values between patients. Kaplan-Meier product limit calculations were used to estimate the occurrence of survival and leukemia-free survival (LFS) [22]. Values are expressed as a percentage with the 95% Rothman interval. Comparison of two such estimates relied on the Log-Rank test. Cumulative incidence was calculated to express the probability of relapse with death from other causes as competing risk [23]. Transplant mortality (TM) was defined as death without evidence of previous relapse. Patients who died post-transplant, after relapse were considered as dying of relapse whatever the ultimate cause of death. Leukemia-free survival (LFS) was calculated with relapse or death as endpoint, whichever occurred first and censoring patients alive and relapse-free at the time of last contact. Analysis was performed on February 15th, 1999 allowing a minimal follow-up of 64 months. Four patients (study group: N = 3; control group: N = 1) were lost to follow-up at 25, 39, 50 and 55 months post-transplantation. None of them had relapsed by that date.

RESULTS

One hundred and thirty patients were included in this study over a 33 month period from January 91 until October 93. At the time of transplant, half of the patients (N = 65) were randomized to receive r-IL-2 or to the control group (N = 65). Median age was 37 ± 13 and the sex ratio (M/F) was 77/43. Seventy-eight (60%) of the patients were treated for AML while fifty two (40%) others suffered from ALL. Eighty-seven patients (67%) had an informative cytogenetic examination at time of diagnosis. Cytogenetic prognosis was classified as favorable for nineteen of the available patients (22%), unfavorable in fifteen (17%) and intermediate for the fifty-three (61%) others. Ten patients (8%) reached CR1 after two courses of induction. The median time between diagnosis and BMT was 4.9 months (range: 2.9-9.7). Both groups were well-balanced for all variables studied. In addition, initial transplant course and hematological recovery was similar in the 2 groups (data not shown).

Thirty-eight (59%) of the 65 patients originally randomized into the study group subsequently started r-IL-2. Eleven of the 27 patients who failed to receive r-IL-2 (41%) (AML: N = 8; ALL: N = 3) did not receive treatment due to insufficient hematological recovery, seven (26%) (AML: N = 4; ALL: N = 3) because of an early relapse and nine (33%) (all AML) for various reasons: poor medical condition (N = 5) and refusal (N = 4). There was no difference between the characteristics of the patients who started r-IL-2 and those who did not. Fewer patients with AML started immunotherapy than in the ALL cohort (p < 0.03) and those who did, did so significantly later (p < 0.02) (Table 2). Overall, patients started their treatment at a median of sixty-eight days (23-140) after transplant and received 77% (16-100) of the scheduled dosage. They received a median of 120 x 10⁶ IU/m² (25-156) over 10 (3-13) days during a total period of 56 (3-78) days. Thirteen patients (AML: N = 9; ALL: N = 4) started each of the five cycles and seven of them received more than 95% of the total scheduled dose (AML: N = 4; ALL: N = 3). Of the 86 cycles of r-IL-2 effectively started, treatment was discontinued on 35 occasions (40%) for one or several reasons. The major reason for suspending treatment was related to the occurrence of a capillary leak syndrome (38%). Other reasons are as following : liver toxicity (14%), thrombopenia (11%), high fever (9%), neurological toxicity (9%), relapse (9%) severe infection (5%) and refusal (5%). None of the patients who received r-IL-2 experienced a secondary graft failure.

The median follow-up was equal to seven years (5.4-8.1 years). Overall, 79 patients relapsed (study group: 43 (66%); control group: 36 (55%): p = Ns) and 81 died. Six patients died without relapsing. Three succumbed to a sepsis during the initial course of the transplant (control group: N = 2, study group: N = 1), two died from r-IL-2 related acute respiratory distress syndrome (ARDS) and finally, one (control group) had a late secondary neoplasia and died from its evolution. The estimates for survival and LFS probability are respectively : 35% (26-45) and 32% (24-42) with no statistical difference between AML and ALL.

With an intent-to-treat analysis, no difference in outcome existed between study and control groups for the whole population (Table 3, Figure 1), or separately for AML or ALL. Patients who started the r-IL-2 treatment had an eventual outcome which was no different from those who did not received r-IL-2.

Specifically, forty-four (56%) patients with AML relapsed (study group : 25 of 45 (63%); control group: 19 of 34 (50%): p = Ns) and thirty-two are still alive, twenty-eight of them leukemia-free. This led to the following estimates of survival and LFS: 38% (24-53) versus 47 (32-62) (p = Ns) and 30% (18-45) versus 36% (19-57) (p = Ns) respectively for patients randomized to study and control group. Of the 19 patients with AML effectively treated with r-IL-2, 13 (68%) relapsed and six are alive, four leukemia-free leading to the following estimates for survival and LFS: 32% (15-54) and 21% (9-43) (Figure 2).

Similarly, of the 52 patients suffering from ALL, 35 (67%) relapsed (study group: 18 of 25 (72%); control group: 17 of 27 (63%): p = ns) and 13 are alive and leukemia-free. Kaplan-Meier estimates for survival and LFS are : 25% (12-46) versus 36% (21-55) (p = NS) and 28% (14-48) versus 37% (22-56) (p = NS) respectively for patients randomized to study and control group. For the same end-points 12 of the 19 patients (63%) treated with r-IL-2 relapsed and 15 are alive and leukemia-free with the following survival and LFS estimates: 33% (15-37) and 37 % (19-59) (Figure 2).

No factors have been found to be predictive of relapse in patients who received r-IL-2.

DISCUSSION

This study represents the first randomized trial which prospectively attempts to assess the impact of r-IL-2 in the treatment of AL, after consolidation of a CR1 with autologous BMT. Characteristics of the patients included in the study represented a common population eligible for autologous BMT in the early 90's. The absence of a different outcome in the two arms during the initial follow-up led us to re-analyze outcome after a longer observation period. This study showed that no benefit was observed after a minimal follow-up of over five years and a median time of observation of over seven years, during which time any impact of r-IL-2 should have manifested itself. These results elicit a number of different comments.

First, we present similar results for both AML and ALL. Although separate analysis did not show any difference, the relatively low number of patients who received r-IL-2 may preclude any conclusion in this regard. However, these diseases are not equivalent. In this field it is particularly interesting to note that there are very few encouraging reports concerning the use of r-IL-2 in ALL [24], which is in sharp contrast with AML [4, 25, 26]. This may warrant further investigation of this concept in the latter disease.

Then, it is obvious that a large fraction (42%) of the 65 patients randomized to receive r-IL-2 therapy did not receive it. This proportion is even higher if considering patients treated for AML (53%). Two reasons contributed to this lack of compliance. Patients were randomized at the time of autologous transplant for treatment which was scheduled to start two months later. This choice was made with the intention of analyzing transplant and immunotherapy as a single therapy. However, this kind of approach maximized the risks for patients not receiving r-IL-2 therapy as scheduled. The reality of this risk was in fact confirmed in this study: two thirds of the patients (N = 18) failed to receive r-IL-2 because either they had a poor hematological recovery or developed an early relapse. Overall, on Day 90 post-transplant, 39 (30%) patients in the study either had relapsed or had not reached adequate counts. Today, this proportion would probably be lower if blood cell transplant was used: this would probably result in a better hematopoietic recovery and facilitate the possibility of starting the r-IL-2 treatment sooner and thus possibly prevent the risk of relapse otherwise observed.

The third problem with this late schedule of an intermediate dose of r-IL-2, was the high degree of toxicity. Only 18% of patients who started r-IL-2 received 95% or more of the scheduled amount. As already reported [11], the occurrence of capillary leak syndrome represented the major reason for discontinuation. These toxicities led directly to the death of two patients. In addition, we previously reported that this schedule was associated with a high immune stimulation [19].

CONCLUSION

One conclusion from this study is that this therapeutic strategy was unable to prevent relapse: relapse, survival and LFS did not statistically differ between the two groups. Under no circumstances can the results be considered as promising. One can argue that this specific schedule may explain the failure, and it is true that this regimen may suffer from imperfections. As the r-IL-2 regimen may have started too late, it may be vital to investigate the possibility of immuno-modulation at an earlier stage, and notably during the first month post-transplant. In the allo setting, we reported the adverse effect on the graft-versus-leukemia effect, when the IL-2 receptor is blocked during the early post-transplant period [27]. Furthermore, the fine tuning of the timing of r-IL-2 administration has been suggested by others [28]. Dose and schedule may also have been inadequate and this may have contributed to the negative results. It is true that the antileukemic effect in the allo setting is a long lasting effect due to the continuous and permanent action exerted by the allogeneic graft. Many facts illustrate this point and notably the lasting positivity of molecular signals after allo BMT for chronic myeloid leukemia without further relapse. However, with regard to r-IL-2 therapy, the optimal ratio between timing, dose and duration remains to be defined.

On the other hand, it can be also be noted that the treatment administered to patients could not be regarded as insignificant considering the observed toxicities. Despite this, we were unable to detect any encouraging effect in the treated patients. This may signify that increased leukemic control may be quite difficult to achieve in the setting of r-IL-2 therapy alone, what-ever the improvement of the schedule.

The present results may encourage further investigations into better target recognition and more potent effector cells. It is possible that r-IL-2 may have been inefficient in this trial because targets did not express a sufficient antigenic signal, which, on the other hand, may be the case in more advanced disease [25, 26]. Extensive work is presently being done to increase the recognition of signals, with the most promising studies being the investigation of the use of dendritic cells [29]. Finally, the role of effector cells has perhaps been underestimated: some investigators have reported using lymphokine-activated cells with interesting results [25]. Other trials were more focused on cell activation and initial reports are also promising [30].

Finally, these reports stress the need for further investigation, to both elucidate these various aspects and to develop more efficient immunotherapy, given the very disappointing results for patients treated with chemotherapy alone.

Acknowledgements. We thank the investigators of other transplant teams for their active participation: C. Auzanneau (Poissy), J.P. Vernant (Créteil); M. Leporrier (Caen); M. Legros (Ý) (Clermont-Ferrand) and C. Gisselbrecht (Hôpital Saint-Louis, Paris). We would also like to express our gratitude to Professor Finn Bo Petersen, University of Utah, USA for his helpful advice.

Support. This trial was supported in part by a grant from the comité du Var de la Ligue Nationale de Lutte contre le Cancer.

REFERENCES

1. Reiffers J, Gaspard M H, Maraninchi D, Michallet M, Marit G, Stoppa A M, Corront B, David B, Gastaut J A, Scotto J J, Broustet A, Carcassonne N, Hollard D. 1989. Comparison of allogeneic or autologous bone marrow transplantation and chemotherapy in patients with acute myeloid leukaemia in first remission: a prospective controlled trial. Br. J. Haematol. 72: 57.

2. Zittoun R A, Mandelli F, Willemze R, de Witte T, Labar B, Resegotti L, Leoni F, Damasio E, Visani G, Papa G, Caronia F, Hayat M, Stryckmans P, Rotoli B, Leoni P, Peetermans M E, Dardenne M, Vegna M L, Petti M C, Solbu G and S S. 1995. Autologous or allogeneic bone marrrow transplantation compared with intensive chemotherapy in acute myelogenous leukemia. N. Engl. J. Med. 332: 217.

3. Cassileth P A, Harrington D P, Appelbaum F R, Lazarus H M, Rowe J M, Paietta E, Willman C, Hurd D D, Bennett J M, Blume K G, Head D R, Wiernik P H. 1998. Chemotherapy compared with autologous or allogeneic bone marrow transplantation in the management of acute myeloid leukemia in first remission. N. Engl. J. Med. 339: 1649.

4. Maraninchi D, Blaise D, Viens P, Brandely M, Olive D, Lopez M, Sainty D, Marit G, Stoppa A M, Reiffers J, Gratecos N, Bertau-Perez P, Mannoni P, Mawas C, Hercend T, Sebahoun G, Carcassonne Y. 1991. High-dose recombinant interleukin-2 and acute myeloid leukemias in relapse. Blood 78: 2182.

5. Maraninchi D, Vey N, Viens P, Stoppa A M, Archimbaud E, Attal M, Baume D, Bouabdallah R, Demeoq F, Fleury J, Michallet M, Olive D, Reiffers J, Sainty D, Tabilio A, Tiberghien P, Brandely M, Hercend T, Blaise D. 1998. A phase II study of interleukin-2 in 49 patients with relapsed or refractory acute leukemia. Leuk. Lymphoma 31: 343.

6. Stoppa A M, Fossat C, Blaise D, Viens P, Brandely M, Pourreau C N, Sainty D, Novakovitch G, Miquel M, Juhan-Vague I, Maraninchi D. 1991. Interleukin-2 induces chemotactic deficiency in patients with onco-hematologic malignancies and autologous bone marrow transplantation. Eur. Cytokine Netw. 2: 231.

7. Fossat C, Sainty D, Stoppa A M, Blaise D, David M, Maraninchi D, Juhan-Vague I. 1993. In vitro inhibition of interleukin-2-induced defective polymorphonuclear chemotaxis by TNF inhibitor. Eur. J. Haematol. 51: 13.

8. Olive D, Lopez M, Blaise D, Viens P, Stoppa A M, Brandely M, Mawas C, Mannoni P, Maraninchi D. 1991. Cell surface expression of ICAM-1 (CD54) and LFA-3 (CD58), two adhesion molecules, is up-regulated on bone marrow leukemic blasts after in vivo administration of high-dose recombinant interleukin-2. J. Immunother. 10: 412.

9. Olive D, Chambost H, Sainty D, Stoppa A M, Blaise D, el Marsafy S, Brandely M, Mannoni P, Mawas C, Maraninchi D. 1994. Modifications of leukemic blast cells induced by in vivo high-dose recombinant interleukin-2. Leukemia 8: 1230.

10. Foa R. 1996. Interleukin-2 in the management of acute leukemia. (Review) Br. J. Haematol. 92: 1.

11. Blaise D, Maraninchi D. 1998. Interleukin-2 in the treatment of acute leukemia. (Review) Leuk. Res. 22: 1165.

12. Meloni G, Foa R, Vignetti M, Guarini A, Fenu S, Tosti S, Tos A G, Mandelli F. 1994. Interleukin-2 may induce prolonged remissions in advanced acute myelogenous leukemia. Blood 84: 2158.

13. Reittie J E, Gottlieb D, Heslop H E, Leger O, Drexler H G, Hazlehurst G, Hoffbrand A V, Prentice H G, Brenner M K. 1989. Endogenously generated activated killer cells circulate after autologous and allogeneic marrow transplantation but not after chemotherapy. Blood 73: 1351.

14. Blaise D, Olive D, Stoppa A M, Viens P, Pourreau C, Lopez M, Attal M, Jasmin C, Monges G, Mawas C, Mannoni P, Palmer P, Franks C, Philip T, Maraninchi D. 1990. Hematologic and immunologic effects of the systemic administration of recombinant interleukin-2 after autologous bone marrow transplantation. Blood 76: 1092.

15. Blaise D, Viens P, Olive D, Stoppa A M, Gabert J, Pourreau C N, Attal M, Gaspard M H, Mannoni P, Jasmin C, Palmer P, Franks C, Michel G, Mawas C, Baume D, Philip T, Maraninchi D. 1991. Recombinant interleukin-2 (r-IL-2) after autologous bone marrow transplantation (BMT): a pilot study in 19 patients. Eur. Cytokine Netw. 2: 121.

16. Blaise D, Stoppa A M, Viens P, Sainty D, Fossat C, Miquel M, Olive D, Bouabdallah R, Gabert J, Baume D, Maraninchi D. 1992. Intensive immunotherapy with recombinant
IL-2 after autologous bone marrow transplantation is associated with a high incidence of bacterial infections [letter]. Bone Marrow Transplant. 10: 193.

17. Costello R, Blaise D, Jacquemier J, Monges G, Stoppa A M, Viens P, Olive D, Bouabdallah M, Brandely M, Gastaut J A. 1995. Induction of cutaneous'graft-versus-host like' reaction by recombinant IL-2 after autologous bone marrow transplantation [letter]. Bone Marrow Transplant. 16: 199.

18. Attal M, Blaise D, Marit G, Payen C, Michallet M, Vernant J P, Sauvage C, Troussard X, Nedellec G, Pico J, Huguet F, Stoppa A M, Broustet A, Sotto J J, Pris J, Maraninchi D, Reiffers J. 1995. Consolidation treatment of adult acute lymphoblastic leukemia: a prospective, randomized trial comparing allogeneic versus autologous bone marrow transplantation and testing the impact of recombinant interleukin-2 after autologous bone marrow transplantation. BGMT Group. Blood 86: 1619.

19. Blaise D, Attal M, Pico J L, Reiffers J, Stoppa A M, Bellanger C, Molina L, Nedellec G, Vernant J P, Legros M, Gabus R, Huguet F, Brandely M, Hercend T, Olive D, Maraninchi D. 1997. The use of a sequential high dose recombinant interleukin-2 regimen after autologous bone marrow transplantation does not improve the disease free survival of patients with acute leukemia transplanted in first complete remission. Leuk. Lymphoma 25: 469.

20. Dastugue N, Payen C, Lafage-Pochitaloff M, Bernard P, Leroux D, Huguet-Rigal F, Stoppa A M, Marit G, Molina L, Michallet M, Maraninchi D, Attal M, Reiffers J. 1995. Prognostic significance of karyotype in de novo adult acute myeloid leukemia. The BGMT group. Leukemia 9: 1491.

21. Faderl S, Kantarjian H M, Talpaz M, Estrov Z. 1998. Clinical significance of cytogenetic abnormalities in adult acute lymphoblastic leukemia. (Review) Blood 91: 3995.

22. Kaplan E, Meier P. 1958. Nonparametric estimation for incomplete observations. J. Am. Stat. Assoc. 53: 457.

23. Gooley T A, Leisenring W, Crowley J, Storer B E. 1999. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. (Review) Stat Med 18: 695.

24. Meloni G, Foa R, Tosti S, Vignetti M, Gavosto F, Mandelli F. 1990. IL-2 in the treatment of chronic myeloid leukemia after lymphoid blast crisis: a pilot study. Haematologica 75: 502.

25. Benyunes M C, Massumoto C, York A, Higuchi C M, Buckner C D, Thompson J A, Petersen F B, Fefer A. 1993. Interleukin-2 with or without lymphokine-activated killer cells as consolidative immunotherapy after autologous bone marrow transplantation for acute myelogenous leukemia. Bone Marrow Transplant 12: 159.

26. Meloni G, Foa R, Vignetti M, Guarini A, Fenu S, Tosti S, Tos A G, Mandelli F. 1994. Interleukin-2 may induce prolonged remissions in advanced acute myelogenous leukemia. Blood 84: 2158.

27. Blaise D, Olive D, Michallet M, Marit G, Leblond V, Maraninchi D. 1995. Impairment of leukaemia-free survival by addition of interleukin-2-receptor antibody to standard graft-versus-host prophylaxis. Lancet 345: 1144.

28. Sykes M, Abraham V S, Harty M W, Pearson D A. 1993. IL-2 reduces graft-versus-host disease and preserves a graft-versus- leukemia effect by selectively inhibiting CD4⁺ T cell activity. J. Immunol. 150: 197.

29. Charbonnier A, Gaugler B, Sainty D, Lafage-Potchitaloff M, Olive D. 1999. Human acute myeloblastic leukemia differentiate in vitro into mature dendritic cells and induce the differentiation of cytotoxic T cells against autologous leukemias. Eur. J. Immunol. In press.

30. Mazumder A.1997. Experimental evidence of interleukin-2 activity in bone marrow transplantation. (Review) Cancer J. Sci. Am. 3: S37.