Forty-six patients with severe aplastic anemia (median age, 23 years) were treated with high-dose cyclophosphamide followed by infusion of marrow from an HLA-identical family member. To evaluate postgrafting prophylaxis for graft-v-host disease (GVHD), they were entered into a prospective randomized trial comparing the effect of a combination of methotrexate and cyclosporine (n = 22) to that of methotrexate alone (n = 24). Forty-four of the forty-six patients had evidence of sustained marrow engraftment. Only one patient in each of the two study groups showed graft rejection. A significant reduction in the cumulative incidence of grades II to IV acute GVHD was seen in patients given methotrexate/cyclosporine (18%) compared with those given methotrexate alone (53%) (P = .012). In three patients given methotrexate alone, grade III developed, and in six, grade IV acute GVHD developed, compared with none given methotrexate/cyclosporine. Eighteen of the 22 patients given methotrexate/cyclosporine and 15 of the 24 given methotrexate alone are alive between 5.5 and 44.5 months (median, 18 months), with actuarial survival rates at 2 years of 82% and 60%, respectively (P = .062). The incidence of fatal infections was higher in patients given methotrexate alone, whereas there are as yet no significant differences in the incidence of chronic GVHD. We conclude that methotrexate/cyclosporine treatment resulted in a significant decrease in the incidence and severity of acute GVHD in patients who received transplants for severe aplastic anemia and thus an improvement in survival.

Platelet transport towards the vessel wall is influenced by the hematocrit, red blood cell (RBC) size, and shape. Recent in vitro studies have indicated that RBC deformability may also influence platelet transport. The observation that isoxsuprine, a known vasodilating drug, caused increased RBC deformability in vitro and decreased platelet transport in vitro prompted us to study the effects of this drug in vivo. The study was performed in a double-blind cross-over study of isoxsuprine v placebo in ten patients with peripheral arterial insufficiency. RBC deformability was estimated from viscosity measurements using the blood viscosity equation of Dintenfass and expressed as T value. Platelet transport was studied in an annular perfusion chamber according to Baumgartner. Human umbilical arteries were used as blood vessels. Perfusion studies were performed with whole blood or with RBCs of the patients mixed with normal platelets and plasma at a standardized hematocrit and platelet count. An increase in RBC deformability concomitant with a decrease in platelet adherence was observed in patients on isoxsuprine with a drop in T value of approximately 0.06 (from 0.91 toward 0.86), and a concomitant decrease in platelet adherence of 20% to 40%. These observations differed significantly from the results in the placebo group and showed a significant group-period interaction at the cross-over of medication (analysis of variance). The effects on platelet adherence were observed at high vessel wall shear rate (1,800 s-1) with perfusates consisting of patients' RBCs and donor plasma and platelets at standardized hematocrit and platelet count. No differences were observed under these conditions at a shear rate of 300 s-1. When whole blood of patients was used, nonsignificant effect was observed at shear rates of 300 s-1 and 1,800 s-1. This was probably caused by the added noise due to variations in hematocrit and platelet number. These data demonstrate that isoxsuprine increases RBC deformability, and they suggest the possibility of decreasing platelet-vessel wall interaction in vivo by manipulation of RBC deformability.

We have recently completed a double-blind, placebo-controlled, noncrossover study, the goal of which was to determine whether cetiedil citrate (cetiedil) could affect the course of vaso-occlusive crises in sickle cell disease. Patients, who presented to the emergency room at least 4 but no more than 24 hours after the onset of a painful vasoocclusive crisis severe enough to require hospitalization, were considered candidates for the study. Each patient received either placebo or cetiedil at one of the following three dosages: 0.2, 0.3, or 0.4 mg/kg body weight. The assigned drug dosage was given as a 30 minute intravenous infusion every 8 hours for 4 consecutive days. A total of 67 patients was enrolled in the study. Cetiedil, at its highest dosage (0.4 mg/kg body weight), was found to be significantly superior to placebo both in reducing the number of painful sites present on all 4 treatment days and in shortening the total time in crisis. No serious adverse reactions were observed during the course of the study. We conclude that cetiedil, given at a dosage of 0.4 mg/kg body weight, is therapeutically advantageous for sickle cell crisis.

The sensitivity of human myeloblastic leukemic (CFU-L) and normal hemopoietic stem cells (CFU-GM and BFU-e) to Asta Z 7557 (INN Mafosfamide) was studied with regard to autologous bone marrow transplantation (ABMT) with cleansed marrow for consolidation therapy in adult patients with acute leukemia (AL) in remission. Establishment of the dose-response curves for CFU-GM (n = 37), BFUe (n = 11), and myeloblastic CFU-L (n = 9) demonstrated a wide range of sensitivity from patient to patient for all three progenitors. Whereas CFU-L, CFU-GM, and BFU-e grown in semisolid cultures disclosed similar sensitivities to Asta Z 7557, long-term culture (LTC) studies (n = 41) indicated a higher resistance of early progenitors. In an effort to achieve a maximum tumor cell kill and yet spare a sufficient amount of normal stem cells to ensure consistent engraftment, we defined the optimal dose for marrow cleansing as the dose sparing 5% CFU-GM (LD95). This dose was established from a preincubation test (PIT) realized on a 10-mL marrow aspirate taken 15 days before marrow collection in each individual patient. Twenty-four adult patients while in remission of AL (20 in complete remission, four in partial remission) were consolidated by cyclophosphamide 60 mg/kg X 2 and total body irradiation at 10 Gy followed by ABMT with marrow cleansed by Asta Z 7557 according to the specification described above. Patients were divided in two groups: group 1, unfavorable prognosis (11 patients); group 2, standard prognosis [13 patients in first complete remission (CR)]. All patients engrafted on leukocytes (median day for recovery to 10(9)/L: day 30), patients with ALL recovered faster than patients with ANL (median day 19 v 34). Similarly, recovery of platelets to 50.10(9)/L occurred sooner in patients with ALL (median day 67, range day 23 through 90) whereas three patients with acute nonlymphoblastic leukemia (ANLL) in group 2 had to be supported with platelet transfusions for more than one year. In group 1, six patients had recurrent tumor within six months; three patients died from toxicity with no evidence of tumor. Two patients are still disease-free with a short follow-up (nine and ten months). In group 2, two patients died from toxicity with no evidence of leukemia three and 16 months post-ABMT. One patient with a M5 ANLL and one patient with ALL relapsed at six and 15 months, respectively. Nine patients have remained in CR or are disease-free with a median follow-up of 22 months.(ABSTRACT TRUNCATED AT 400 WORDS)

Sixteen patients with leukemia in relapse or second to third remission, 5 to 27 years old (median, 17), were given cyclophosphamide (60 mg/kg X 2) and total body irradiation (2.25 Gy for each of seven days) followed by unmodified marrow grafts from HLA-identical siblings. Patients did not receive posttransplant immunosuppression and were followed a median of nine months (range, 5-17). Prompt engraftment was sustained in 12 patients with a median time of 16 days (range, 10 to 63) to achieve 500 neutrophils/mm3. One patient failed to engraft, one had delayed engraftment, and two had late poor graft function. All 15 with engraftment developed moderate to life-threatening graft-v-host disease (GVHD, eight grade II and seven grade III-IV). This syndrome was hyperacute (median onset eight days [range, 7 to 29] posttransplant) and manifest by severe skin disease (14 patients at stage 3 and one at stage 4), fever (ten patients), and liver (four patients, stage 3-4) or gut (four patients, stage 3-4) involvement. Serial tissue biopsies confirmed acute GVHD in 13 of 15 patients. Ten were treated with antithymocyte globulin and cyclosporine (four survive), and four with corticosteroids (two survive). Actuarial survival to 17 months was 37%. Causes of death included interstitial pneumonia (four), infection (three), graft failure (one), venocclusive disease (one), and relapse of leukemia (one). Age-matched controls receiving standard methotrexate after transplant had comparable relapse-free survival but only a 25% incidence of grade II-IV acute GVHD (P less than .0001). We conclude that deleting posttransplant immunosuppression is associated with frequent and severe hyperacute GVHD, infectious complications, and occasional poor graft function.

To determine the effects of the "maturation-inducing" agents 13-cis retinoic acid and 1,25 dihydroxyvitamin D3 on marrow cells from normal individuals and patients with myelodysplastic syndromes (MDS), we assessed marrow hemopoietic clonogenicity and differentiation response patterns to these agents. These vitamins caused increased proliferation in vitro of normal clonogenic marrow myeloid precursor cells (CFU-GM), decreased erythroid precursors (BFU-E), and no change in multipotent stem cells (CFU-GEMM). Marrow hemopoietic colony-forming cell incidence was generally subnormal in the 22 MDS patients evaluated. In vitro exposure to both agents caused various patterns of alteration of MDS hemopoietic colony and cluster formation, with similar but more pronounced effects evoked by retinoic acid. In the vast majority of MDS patients, enhanced marrow clonal granulocyte-monocyte differentiation and decreased BFU-E growth were noted after in vitro exposure to these vitamins. Correlation of biological effects was demonstrated between in vivo changes of peripheral neutrophil counts and in vitro responses of myeloid precursors for ten MDS patients treated with an eight-week therapeutic course of retinoic acid. Cytogenetic analyses indicated persisting aneuploidy or coexisting normal and aneuploid karyotypes in the cultured MDS myeloid cells and (with one exception) in native marrow cells from the treated patients. The varying responses of the MDS cells may monitor differing proportions of normal versus leukemic marrow cells susceptible to proliferative and differentiative expression on exposure to these agents.

The synthetic vasopressin derivative desmopressin (DDAVP) shortens a prolonged bleeding time (BT) in patients with uremia, congenital platelet dysfunction, and von Willebrand disease. To establish the limits of the clinical usefulness of DDAVP, a controlled randomized study was carried out in 53 patients and ten volunteers with different conditions that have in common a prolonged BT. DDAVP significantly shortened the BT in 21 cirrhotics (P less than .01), in eight patients with unclassified prolonged BT (P less than .05) and in ten volunteers taking the antiplatelet drugs aspirin (P less than .05) and ticlopidine. The BT changes were not statistically significant in 15 patients with severe thrombocytopenia nor in nine with congenital platelet dysfunction, even though a few patients with storage pool deficiency responded with a marked BT shortening. Our findings indicate that DDAVP might be given when biopsies or other surgical procedures must be carried out in patients with prolonged BT. However, the compound is often ineffective in patients with thrombocytopenia or congenital platelet dysfunction.

Forty-eight patients with chronic myelocytic leukemia, aged 11 to 47, were treated with high-dose cyclophosphamide and fractionated total body irradiation, followed by infusion of marrow from HLA-identical siblings. They were randomized to receive either methotrexate (MTX) (n = 23) or cyclosporine (CSP) (n = 25) as postgrafting prophylaxis for graft-v-host disease (GVHD). All patients had evidence of sustained hematopoietic engraftment. Seventeen of the 25 patients receiving CSP and 17 of the 23 patients receiving MTX are alive between one and almost four (median, 1.7) years, with an actuarial survival rate at three years of 62% and 66%, respectively (P = .60). Also, with respect to most other parameters studied, the two drugs were identical. The probability of acute GVHD was .42 and .46, respectively (P = .70), that of chronic GVHD, .50 and .63 (P = .44), and that of death from transplant-related causes, .30 and .24 (P = .51). There were no differences in the speed of granulocyte and platelet engraftment (P = .82 and .94, respectively), and the duration of hospitalization was comparable (P = .58). Patients receiving MTX required red cell transfusions for a shorter period of time (P = .02), but had a slightly increased morbidity from early oral mucositis. The leukemia recurrence rates were comparable (P = .60). With the regimens used in this study, we conclude that CSP failed to reduce the incidence of GVHD and improve the survival of patients with chronic myelocytic leukemia when compared to results with standard MTX.

We analyzed the effect of antithymocyte globulin (ATG) with or without androgens in 121 patients with aplastic anemia. Fifty-three patients with moderate to severe aplastic anemia were prospectively randomized to receive ATG with or without oral androgens. Eleven of 26 patients (42%) receiving ATG plus androgen responded, including three complete and eight partial responses. Twelve of 27 patients (44%) receiving ATG plus placebo responded, including five complete and seven partial responses. The difference in response rates was not significant (P greater than .9). Survival was also comparable in the two groups; for patients with severe aplastic anemia, actuarial survival at two years was 55% +/- 24% (95% confidence interval) in patients receiving ATG plus androgen compared with 50% +/- 24% in the ATG plus placebo group (P = .65). Furthermore, results in both groups were indistinguishable from those obtained in 68 historical controls receiving ATG without androgens. These data indicate that androgens are not required in order to respond to antithymocyte globulin and the addition of androgens, as used in this trial, did not significantly improve response rates to ATG treatment.

This phase I study was conducted to determine the maximal safe concentration of 4-hydroperoxycyclophosphamide (4HC) that could be used for in vitro treatment of bone marrow from patients with acute leukemia undergoing autologous bone marrow transplantation. Concentrations of 40 to 120 micrograms/mL of 4HC were used in 30 patients with relapsed or high-risk acute leukemia and in six patients with nonleukemic malignancies. All patients received marrow-lethal cytoreductive therapy followed by infusion of the 4HC-treated marrow. Complete inhibition of granulocyte and macrophage colony-forming cells was obtained at 80 micrograms/mL. Nevertheless, only one transplant-related death and otherwise full hematologic recovery was observed at concentrations of 4HC up to 100 micrograms/mL. At 120 micrograms/mL, there were three transplant-related deaths, including two of the three patients who required the infusion of reserve marrow. Among the acute leukemia patients, three remain in complete remission at 1,337, 1,017, and 967 days after transplant. Among the nonleukemic patients, two remain in complete remission at 1,081 and 1,017 days after transplant. At the maximum safe concentration of 4HC (100 micrograms/mL), satisfactory hematologic recovery can be obtained, despite elimination of detectable hematopoietic progenitors.

Eleven patients with B lymphocytic malignancy were treated with mouse monoclonal anti-idiotype antibodies. All but one of the patients in this study had received extensive prior treatment with conventional lymphoma therapy. All antibodies were prepared against, and uniquely reactive with, the patient's own tumor. Ten patients were treated with a single antibody, but one patient received three antibodies concurrently. The treatment protocol initially used an escalating dose schedule that was intended to evaluate toxicity, pharmacokinetics and, eventually, to achieve appreciable levels of free mouse antibody in the circulation. The last two patients received substantial initial doses. Tumor sampling was performed before and during therapy to evaluate tissue penetration by antibody. None of the patients had serum paraproteins by routine clinical testing, but six had idiotype protein detectable by a sensitive immunoassay at levels greater than 1 microgram/mL, two of which were greater than 200 micrograms/mL. Plasmapheresis was capable of reducing these levels temporarily. However, the presence of serum idiotype increased the requirement for mouse antibody to achieve tumor penetration. Another obstacle to treatment was immune response to mouse Ig, which occurred in five of the 11 patients. Once an immune response had begun, further infusions of antibody were not capable of reaching the tumor or inducing tumor regression and were associated with toxicity. Our initial patient remains in an unmaintained complete remission 42 months after receiving antibody. Five of ten additional patients have had objective remissions that were also clinically significant. However, these remissions were not complete and were of relatively short duration. This therapy shows promise as an alternative modality for the treatment of B cell malignancy. Further study will be needed to determine the mechanisms of the antitumor effect and to improve the clinical results.

Seventy-five patients, 13 to 49 years of age, with acute nonlymphoblastic leukemia in first remission were treated with cyclophosphamide, fractionated total body irradiation, and marrow transplantation from an HLA-identical sibling and randomized to receive either cyclosporine (CSP) (n = 36) or methotrexate (MTX) (n = 39) as prophylaxis for graft-v-host disease (GVHD). All patients engrafted, and 22 who were given CSP and 21 who were given MTX, are alive at 20 to 47 (median, 35) months (P = .5). Engraftment as assessed by granulocyte recovery (P less than .0005) and platelet transfusion requirement (P = .01) was faster in patients on CSP. Twelve patients (33%) on CSP and 22 (56%) on MTX developed acute GVHD of grades II through IV (P = .07) and 15 of 30 on CSP and 14 of 32 on MTX that were at risk developed chronic GVHD. The most frequent causes of death were interstitial pneumonitis and marrow relapse of leukemia, which occurred with similar frequency in both groups. Beneficial effects observed in patients on CSP included less severe mucositis and shorter duration of hospitalization; adverse effects included renal function impairment and hypertension. These data confirm that CSP is a useful immunosuppressant in patients undergoing marrow transplantation but fail to show a significant improvement in survival as compared with the standard regimen of MTX.

Twenty-two patients with hairy cell leukemia were treated with biosynthetic (recombinant) alpha-2-interferon in an open-label, single-arm efficacy study. Patients received 2 X 10(6) U/m2 recombinant alpha-2-interferon three times weekly. Therapy was well tolerated subjectively with minimal short-term hematologic toxicity. Two patients had bacterial infections during the period of study, and one patient experienced a short-lived readily reversible rejection of a corneal transplant. Statistical comparison of the mean hematologic indices at study entry and after three to six months of therapy with recombinant alpha-2-interferon indicates a significant improvement in hemoglobin, granulocyte, and platelet counts. Bone marrow biopsies in six of 14 patients after six months of therapy showed a greater than 50% decrease in the infiltration of leukemia cells. We conclude that recombinant alpha-2-interferon is highly effective therapy for hairy cell leukemia.

One hundred fifty-one children with acute myelogenous leukemia (AML) entered the cooperative study BFM-78 between December 1978 and October 1982. The initial therapy consisted of an intensive induction and consolidation regimen over eight weeks with seven different drugs and cranial irradiation. It was followed by maintenance with thioguanine and cytosine arabinoside for two years and additional Adriamycin during the first year. One hundred nineteen (79%) patients achieved a complete remission. Thirteen (9%) children died of early hemorrhages. After a median follow-up time of 36 (12 to 57) months, 47 relapses have occurred, with CNS involvement in seven cases. The life table analysis revealed a probability for overall survival after almost five years of 45% (SD, 4%), for event-free survival 41% (SD, 4%), and for the event-free interval 52% (SD, 5%). Up to now, no relapse was seen after 2 1/2 years. Risk factor analysis showed that early fatal hemorrhages occurred predominantly in children with M5 FAB type and with initial leukocytosis. An initial high WBC count and liver enlargement were unfavorable parameters for achieving remission. No factors could be identified concerning the risk for relapse. These data indicate that the applied treatment strategy is successful in inducing complete remissions in about three fourths of children with AML and also in enhancing considerably the chances for long-term remission.

In order to analyze the usefulness of different types of treatment in relation to the interval since the onset of idiopathic thrombocytopenic purpura (ITP), a collaborative study of 934 adult patients was undertaken. Prednisone was administered to 818 patients, and 32% of them achieved prolonged complete remission (PCR). However, only 14% of patients who had ITP for more than six months achieved a prednisone-induced PCR (P less than .01). Splenectomy was done in 399 patients, and 65% of them achieved PCR; the remission rate did not vary with the interval since the onset of ITP. Of 120 patients with chronic ITP that was refractory to corticosteroids and splenectomy, 91 received either azathioprine or cyclophosphamide; 21% of them achieved PCR and 55% had a favorable response. None of 19 patients treated with vincristine and only one of ten patients treated with vinblastine-loaded platelets achieved PCR.

Previous reports have suggested that the technique of premature chromosome condensation (PCC) is useful for predicting relapse in patients with acute leukemia. However, these studies involved patients had been in complete remission (CR) for various periods of time and had heterogeneous expectations for relapse. The purpose of this study was to further determine the value of PCC in predicting relapse by examining the PCC characteristics of bone marrow specimens from patients with acute leukemia on a common therapeutic regimen after similar periods in CR. The remission durations after the PCC determinations were compared between patients with high or low proliferative potential indices (PPI, or the fraction of G1 cells in late G1 phase). Of 60 patients studied between two and eight weeks after achieving CR, 14 of the 16 patients exhibiting high PPI values (greater than or equal to 35) have relapsed. The mean time from PCC measurement to relapse was 23 weeks. In contrast, only 19 of the 44 patients exhibiting low PPI values have relapsed, with an estimated mean time to relapse of 68+ weeks. Likewise, of 38 patients studied between nine and 15 weeks of CR, nine of the ten patients exhibiting high PPI values have relapsed (mean time to relapse, 23 weeks), while only 16 of 28 patients with low PPI values have relapsed (estimated mean time to relapse, 54+ weeks). The predictive value of the PCC technique was found to be independent of other prognostic factors for the duration of CR, and it identified those patients within the poor prognostic category with a high likelihood of imminent relapse. While similar trends were observed at later time intervals in CR, the differences in relapse rate between patients with high or low PPI values is not significant. These results confirm the usefulness of the PCC technique in predicting relapse in acute leukemia and could aid in the identification of patients who might benefit by an alteration of therapeutic strategy.

One hundred seventy adult patients with acute lymphoblastic leukemia (ALL) or acute undifferentiated leukemia (AUL) were entered into a prospective multicenter therapy trial at 25 hospitals. The aim of the trial was to improve remission duration by using a modified form of an intensified induction regimen that was successful in childhood ALL, to define immunologic subtypes of ALL by use of cell-surface markers, and to extract other possible prognostic factors. The overall complete remission rate was 77.8%. The median overall survival time was 26 months, being 4 months for nonresponders and 32 months for responders. The median remission duration for the 126 patients with complete remission was 20 months. Prognostically favorable factors for remission duration were response to chemotherapy within 4 weeks, age less than 35 years, a low initial leukocyte count, and the immunologic subtypes c-ALL with early response to therapy and T-ALL, where 61% and 58%, respectively, are still in complete remission at 3 years. An adverse influence on remission duration was observed for the subtype null-ALL, with a median survival of 13 months, and for patients with a delayed response to induction therapy, independent of phenotype.

The efficacy of the addition of intensive therapy with daunorubicin (45 mg/m2 IV on days 1, 2, 3) to an otherwise identical induction program consisting of vincristine, prednisone, and L-asparaginase was assessed in 177 previously untreated adults (greater than or equal to 20 years of age) with acute lymphocytic leukemia (ALL). In the prospectively randomized phase of the investigation, 46 patients received daunorubicin in induction, whereas 53 did not. The two groups were otherwise comparable for pretreatment variables. A complete response was observed in 38/46 patients (83%) treated with daunorubicin, compared to 25/53 (47%) induced with vincristine, prednisone, and L-asparaginase alone (P = .003). The high response rate attributable to the use of the anthracycline was confirmed by the nonrandomized treatment of 78 subsequent patients, in whom a complete response rate of 76% was attained. A common program for central nervous system therapy and for maintenance therapy was employed in 103 patients achieving complete response. Maintenance consisted of cycles of 6-mercaptopurine (6-MP) and methotrexate with periodic reinforcement with vincristine and prednisone. Maintenance therapy proved to be minimally toxic. The average duration of complete response was 15 months and was not affected by the induction program employed. Approximately 25% of responders are projected to remain in continuing complete response for 36 months. The failure of the daunorubicin-containing programs to produce a higher percentage of long-term survivors, despite the higher complete response rates achieved, was thought to be due to the use of a maintenance program that was weak in intensity and dependent on reinforcement with vincristine and prednisone. These data clearly establish the increased effectiveness of vincristine, prednisone, L-asparaginase, and daunorubicin, as compared to this combination without daunorubicin, in the induction of complete response in adults with ALL. The results support the concept of an intensive, rather than a conservative, chemotherapeutic approach as the most appropriate strategy for the treatment of adult ALL.

Morphological characteristics of tumor cells have been employed in the prognosis of lymphomas and solid tumors. This report documents an attempt to predict survival from the known cytologic heterogeneity in multiple myeloma. Myeloma cells in bone marrow smears from patients at diagnosis were evaluated by assigning them to morphologically defined categories. Cox's multivariate regression model for censored survival data was used to generate optimal weights, which served as coefficients in two regression equations to estimate death risk from cellular morphology. Step-wise procedures excluded redundant parameters. "Myeloma morphology score" (MMS) discriminates significantly (p less than 0.0001) among 3 stages, with median survival times of 42.5, 30.7, and 9.1 mo. For clinical routine application, "myeloma progression score" (MPS), the weight sum of the proportion of plasmablasts and the extent of bone marrow plasma cell infiltration, is suggested as a simple prognostic tool. Its discriminative power is very high [p less than 10(-9)]. Median survival times of greater than 71.5, 23.4, and 6.1 mo were found for good, moderate, and poor risk groups, respectively. However, staging is not confined to three subgroups, grouping is flexible, and pairs of data can be matched. This fact may prove to be valuable in designing prognosis-controlled clinical trials or theoretical studies on cellular differentiation. Preliminary results suggest changes in morphology due to disease progression and/or the effect of therapy on tumor kinetics. Most importantly, staging according to MPS or MMS may facilitate the adaption of therapy to the current state of the disease in patients with multiple myeloma.

The Southeastern Cancer Study Group conducted a post-remission induction randomized trial in adult acute myelogenous leukemia to assess the efficacy of alternate drug therapy during consolidation and of immunotherapy during maintenance. Of 508 evaluable patients entered into the study, 335 (66%) achieved a complete remission treated with a 7-day infusion of cytosine arabinoside at a dose of 100 mg/sq m/day and 3 days of daunorubicin at a dose of 45 mg/sq m/day. Those in remission were randomized to receive 3 courses of 1 of 3 consolidation regimens: (A) a continuous infusion of 5-azacytidine, 150 mg/sq m/day for 5 days; (B) 5-azacytidine plus beta-deoxythioguanosine, 300 mg/sq m/day for 5 days; or (C) cytosine arabinoside, 100 mg/sq m/day intravenously, and thioguanine, 100 mg/sq m orally every 12 hr, plus daunorubicin, 10 mg/sq m every 24 hr daily for 5 days. There was no difference in relapse rate among the 3 arms. Those completing consolidation and remaining in remission were randomized to 1 of 3 maintenance regimens: (D) chemotherapy, 5-day infusion of cytosine arabinoside and 2 days of daunorubicin (same doses as induction) given every 13 wk for 1 yr; (E) BCG given twice weekly for 1 mo and then monthly for 1 yr; or (F) the combination of regimens D and E. The median duration of remission was significantly better on regimen D (17.4 versus 9.4 and 9.5 mo), and median survival was 29 mo compared to 21 mo for the other regimens. Those given different drugs during consolidation than used for induction (regimens A and B) and subsequent chemotherapy for maintenance (regimen D) had the longest remission durations and survival. Immunotherapy was not as good as intensive chemotherapy for maintenance.