Inhibition of sterol and phospholipid synthesis in HL-60 promyelocytic leukemia cells by inducers of myeloid differentiation.

Myeloid differentiation is induced in HL-60 promyelocytic leukemia cells by dimethyl sulfoxide, retinoic acid, hypoxanthine, and a number of other chemical agents. We questioned whether the induction process was associated with changes in lipid synthesis. With [14C]acetate, a precursor for all cell lipids, a decrease in sterol and phospholipid synthesis (but not triglyceride synthesis) was observed within the first 5 hr after exposure to inducer, a time prior to inhibition of DNA synthesis or cessation of cell growth. Similarly, the membrane fraction of HL-60 cells exhibited decreased incorporation of newly synthesized lipid. Synthesis of phosphatidylcholine from choline as well as from the transmethylation of membrane phosphatidylethanolamine was also inhibited by myeloid inducers. In contrast, neither sterol nor phospholipid degradation was stimulated under these conditions. Both cholesterol and lanosterol were synthesized by growing HL-60 cells, but cholesterol esters were not. Synthesis of sterols was subject to feedback inhibition by cholesterol in the medium, but such feedback inhibition did not affect differentiation in the presence of myeloid inducers and did not alter the effect of myeloid inducers on phospholipid synthesis. Removal of dimethyl sulfoxide at 16 hr permitted a return to normal lipid synthesis and prevented differentiation, whereas removal of dimethyl sulfoxide at 40 hr was followed by continued inhibition of lipid synthesis and progressive differentiation. These studies demonstrate that the induction of myeloid differentiation is associated with an early inhibition of the synthesis of those lipids which are normally a part of cell membranes.

Differential effects of phorbol esters on normal myeloid precursors and leukemic cells: basis for autologous bone marrow reconstitution in acute nonlymphocytic leukemia using phorbol ester-treated bone marrow from patients in remission.

The phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) induces macrophage-like differentiation of HL60 cells and cells from patients with acute nonlymphocytic leukemia (ANLL). We assessed the use of TPA as a means of eradicating residual leukemia from remission bone marrow prior to autologous bone marrow reconstitution. A 30-min incubation with TPA led to marked growth arrest in HL60 cells and in cells from most patients with acute myelogenous leukemia and acute myelomonocytic leukemia, whereas cells from most patients with acute promyelocytic leukemia and acute undifferentiated leukemia demonstrated a lesser degree of growth arrest. Freezing and thawing, a necessary step in autologous reconstitution, had no effect on the cessation of proliferation induced in HL60 or ANLL cells preincubated with TPA for 30 min. Virtually normal myeloid precursor growth occurred in normal or remission bone marrow cells preincubated with TPA and then frozen and thawed. Based on these observations, two patients with advanced ANLL in remission underwent marrow ablative therapy followed by autologous reconstitution using TPA-treated bone marrow. Limited normal hematopoiesis was reestablished in both patients, although they subsequently experienced leukemic relapse. These studies demonstrate that in ANLL cells, TPA stimulates growth arrest; in contrast, hematopoiesis is able to proceed both in vitro and in vivo.

High peripheral blast count in adult acute myelogenous leukemia is a primary risk factor for CNS leukemia.

The lengthening remission duration achievable in acute myelogenous leukemia (AML) places patients at risks for CNS leukemic relapse. We reviewed the data on two Eastern Cooperative Oncology Group (ECOG) trials in acute nonlymphocytic leukemia to determine the incidence of CNS leukemia (CNSL). The incidence of CNSL was 5% (30 of 569 patients) overall, and 3% (ten of 331) in patients in complete remission (CR). A number of factors were evaluated for association with increased risk of CNSL. Men more frequently developed CNSL than women at a three to one ratio, and median presenting WBC counts were higher in affected than unaffected patients (44,200/microL v 17,000/microL, P = .01). The low incidence of CNSL in AML supports the view that CNS prophylaxis is unnecessary. However, because 68% of patients (13 of 19) who developed CNSL early in the course of disease had presenting WBC counts greater than 40,000/microL, screening lumbar punctures should be routinely obtained during induction therapy in patients presenting with high circulating blast cell counts.

A critical comparison of allogeneic bone marrow transplantation and conventional chemotherapy as treatment for acute nonlymphocytic leukemia.

Published data from two centers conducting bone marrow transplantation on patients with acute nonlymphocytic leukemia in first remission were pooled and compared with results from an Eastern Cooperative Oncology Group (ECOG) study in which patients were treated with conventional chemotherapy. A series of adjustments were made to the ECOG sample to account for selection factors that restrict access of patients to transplantation. The transplant sample exhibits considerably higher disease-free survival when compared to the adjusted ECOG series (53% versus 21% at three years). The transplant series is somewhat younger than the ECOG series (median, 24 years versus 28 years). The impact of age on the disease-free survival results is difficult to assess because of the relatively small samples in the different age groups. However, by defining a suitable control group, methodology for making a critical comparison between the two modalities is presented which, if applied to larger samples of patients, should help to resolve the issue. In the absence of data from a large, prospective randomized study, a critical retrospective comparison of available data is essential in the assessment of treatment options.

Identification of early relapsing patients with adult acute nonlymphocytic leukemia by bone marrow biopsy after initial induction chemotherapy.

Bone marrow biopsies obtained from 69 adult patients with acute nonlymphocytic leukemia (ANLL) six to 10 days after initial induction chemotherapy were reviewed blindly to detect the presence of residual leukemia. Discrimination between the presence or absence of leukemic cells was provided by assessment of the numbers, clustering, and nuclear morphology of blasts and promyelocytes. Twenty-six patients had frank leukemia, 25 had no apparent leukemic cells, and 18 had focal residual leukemia. Of 25 patients whose bone marrow contained no detectable residual leukemic cells, 21 gained complete remission without further chemotherapy. These patients had a median duration of remission of 278 days, with five patients still remaining in remission for 578-882 days. Similarly, all of the 18 patients who had focal residual leukemia achieved complete remission without additional chemotherapy; however, all have relapsed with a median duration of remission of 163 days. This study indicates that patients with foci of residual leukemia in their one-week posttreatment bone marrow samples readily achieve remission, but carry a substantial leukemic burden that increases the likelihood of early relapse.

Invasive pulmonary aspergillosis in adult acute leukemia: clinical clues to its diagnosis.

Invasive pulmonary aspergillosis, a leading cause of morbidity and mortality in patients with acute leukemia, is difficult to diagnose antemortem because its signs and symptoms are ill-defined. To refine the clinical description of this infection, we reviewed our experience with 15 pathologically documented cases of invasive pulmonary aspergillosis in a population of 60 patients treated for acute leukemia. Findings occurring significantly more often (P less than or equal to .001) among cases than controls included pleuritic chest pain; acute sinus tenderness, and nasal discharge, epistaxis and eschar; rales; development of multilobar infiltrates after the 14th hospital day; and presence of nodular or cavitary infiltrates. In addition, patients with invasive pulmonary aspergillosis had a significantly prolonged duration of granulocytopenia, more febrile days and febrile episodes without a fever diagnosis and more febrile days on antibiotics (P less than or equal to .001 in all). This complex of findings should improve the clinician's ability to diagnose invasive pulmonary aspergillosis in patients with acute leukemia.

Maintenance chemotherapy prolongs remission duration in adult acute nonlymphocytic leukemia.

The value of maintenance therapy after the achievement of complete remission in adult acute nonlymphocytic leukemia (ANLL) has never been clearly established. A randomized Eastern Cooperative Oncology Group (ECOG) study of postremission therapy compared outcomes in patients who received no further therapy to those administered long-term maintenance chemotherapy. Adverse results in the group administered no further therapy led to early termination of this trial after only 51 patients were randomized. Patients receiving no postremission therapy experienced significantly inferior remission durations (P = .002) compared with patients receiving maintenance therapy. All 26 patients in the group administered no postremission therapy have relapsed, with a median duration of remission of 4.1 months. In contrast, four of 25 patients (16%) who received maintenance therapy remain disease free, with a median duration of remission of 8.1 months.

Pentostatin induces durable remissions in hairy cell leukemia.

Fifty patients with hairy cell leukemia were treated with pentostatin (2'-deoxycoformycin; dCF) for a median of 3 months; 32 (64%) patients achieved complete remission (CR), and 10 (20%) patients achieved partial remission (PR), for an overall response rate of 84%. After reaching maximal response, no maintenance therapy was administered. The median duration of follow-up is now 39 months, and only four of 32 patients in CR and two of 10 patients in PR have relapsed. dCF therapy produces durable long-term, disease-free survival in patients with hairy cell leukemia.

Autologous bone marrow transplant in acute myeloid leukemia in first remission.

PURPOSE: The Eastern Cooperative Oncology Group conducted a prospective study of postremission high-dose chemotherapy and autologous bone marrow transplantation (autoBMT) in a group of uniformly treated adults with de novo acute myeloid leukemia (AML) to evaluate whether intensive, myeloablative therapy in first complete remission (CR) could improve the disease-free survival. PATIENTS AND METHODS: After initial CR was induced by the combination of daunorubicin, cytarabine, and thioguanine, patients not eligible for allogeneic bone marrow transplantation (alloBMT) were offered autoBMT. Within a median of 2 months after CR, and without intervening postremission therapy, bone marrow was obtained, purged by exposure to 4-hydroperoxycyclophosphamide (4-HC), and cryopreserved. High-dose therapy consisted of oral busulfan over 4 days (16 mg/kg total) followed by intravenous (IV) cyclophosphamide 50 mg/kg daily for 4 days. The cryopreserved marrow was then reinfused. RESULTS: Of the 39 patients scheduled for autoBMT, four relapsed before transplantation. Two of the 35 (6%) transplant patients died of transplant-related complications, and 11 (33%) relapsed a median of 8 months after marrow reinfusion. No relapse has occurred after 24 months posttransplant. With a median follow-up of 31 months, the median disease-free survival period for all 39 patients has not been reached; however, 54% +/- 16% of patients are projected to be alive and disease-free at 3 years. CONCLUSION: Long-term, disease-free survival after autoBMT in AML seems to be better than the outcome after conventional-dose postremission therapy and rivals the results of alloBMT.

Randomized comparison of pentostatin versus interferon alfa-2a in previously untreated patients with hairy cell leukemia: an intergroup study.

PURPOSE: Therapy of hairy cell leukemia has markedly improved. Interferon alfa-2a and pentostatin are active agents. The National Cancer Institute organized an intergroup trial to compare these agents prospectively in untreated patients. METHODS: Patients were randomized to receive either interferon alfa-2a (3 x 10(6) U subcutaneously three times per week) or pentostatin (4 mg/m2 intravenously every 2 weeks). Patients who did not respond to initial treatment were crossed over. RESULTS: Of 356 patients on study, 313 were eligible. Among interferon patients, 17 of 159 (11%) achieved a confirmed complete remission and 60 of 159 (38%) had a confirmed complete or partial remission. Among pentostatin patients, 117 of 154 (76%) achieved a confirmed complete remission and 121 of 154 (79%) had a confirmed complete or partial remission. Additional patients achieved criteria for complete remission, but lacked confirmatory follow-up evaluation. Response rates were significantly higher (P < .0001) and relapse-free survival was significantly longer with pentostatin than interferon (P < .0001). The median follow-up duration is 57 months (range, 19 to 82). Myelosuppression was more frequent with pentostatin (P = .013). A multivariate logistic regression analysis of the confirmed complete remissions on pentostatin showed the following factors to be important for achieving a complete remission: high hemoglobin level (two-tailed P = .024), young age (P = .0085), and no or little splenomegaly (P = .0029). CONCLUSION: Both agents were well tolerated. Pentostatin produced higher response rates, and the responses were durable. Patient age and clinical status had an impact on outcome with pentostatin. Pentostatin is effective therapy for hairy cell leukemia.

Phase I study of fludarabine plus cyclophosphamide in patients with previously untreated low-grade lymphoma: results and and long-term follow-up--a report from the Eastern Cooperative Oncology Group.

PURPOSE: To determine the toxicity and recommended phase II doses of the combination of fludarabine plus cyclophosphamide in chemotherapy-naive patients with low-grade lymphoma. PATIENTS AND METHODS: Previously untreated patients with low-grade lymphoma were entered onto dosing cohorts of four patients each. The cyclophosphamide dose, given on day 1, was increased from 600 to 1, 000 mg/m(2). Fludarabine 20 mg/m(2) was administered on days 1 through 5. The first eight patients were treated every 21 days; later patients were treated every 28 days. Prophylactic antibiotics were required. RESULTS: Prolonged cytopenia and pulmonary toxicity each occurred in three of eight patients treated every 3 weeks. The 19 patients treated every 28 days, who were given granulocyte colony-stimulating factor as indicated, did not have undue nonhematologic toxicity. Dose-limiting toxicity was hematologic. At the recommended phase II/III dose (cyclophosphamide 1,000 mg/m(2)), grade 4 neutropenia was observed in 17% of all cycles and 31% of first cycles. Grade 3 or 4 thrombocytopenia was seen in only 1% of all cycles. The median number of cycles per patient was six (range, two to 11) for all patients enrolled. The response rate was 100% of 27 patients entered; 89% achieved a complete and 11% a partial response. Nineteen of 22 patients with bone marrow involvement had clearing of the marrow. Median duration of follow-up was more than 5 years; median overall and disease-free survival times have not been reached. Kaplan-Meier estimated 5-year overall survival and disease-free survival rates were 66% and 53%, respectively. CONCLUSION: The recommended dosing for this combination in patients with previously untreated low-grade lymphoma is cyclophosphamide 1, 000 mg/m(2) day 1 and fludarabine 20 mg/m(2) days 1 through 5. The regimen has a high level of activity, with prolonged complete remissions providing 5-year overall and disease-free survival rates as high as those reported for other therapeutic approaches in untreated patients.

Cerebellar toxicity with high-dose cytosine arabinoside.

CNS dysfunction, especially impaired cerebellar function, is the dose-limiting toxicity associated with high-dose cytosine arabinoside, which precludes doses of greater than 48 g/m2. Four hundred eighteen patients between the ages of 2 and 74 years with leukemia or lymphoma received 36 to 48 g/m2 cytosine arabinoside either alone or with anthracycline antibiotics, 4'-(9-acridinylamino) methane sulfon-m-anisidine (m-AMSA), or total body irradiation. In only 35 of 418 patients (8%) did severe cerebellar toxicity develop; it was irreversible or fatal in four (1%) patients. The age of the patient was a critical factor in the incidence of severe cerebellar toxicity. Patients greater than 50 years old had a statistically significant greater incidence of cerebellar toxicity compared with younger patients (26/137, 19%, v 9/281, 3%; P less than .0005, chi 2). Neither the diagnosis, disease status, sex, nor the regimen altered the incidence of severe cerebellar toxicity (when corrected for age). A second course of high-dose cytosine arabinoside, administered to 62 patients, did not increase the incidence of severe cerebellar toxicity, which occurred in five (8%) of these patients. Two of the five patients had severe toxicity with the initial course. Of the 60 patients with no antecedent cerebellar dysfunction, three (5%) had severe toxicity with the second course: one of 41 patients were less than 50 years old; two of 19 patients were greater than or equal to 50 years. Since the occurrence of severe cerebellar dysfunction is greatly affected by age, reduced doses of high-dose cytosine arabinoside should be given to patients greater than 50 years old, and methods for reducing the cerebellar toxicity should be investigated in these patients.

Activity of fludarabine in previously treated non-Hodgkin's low-grade lymphoma: results of an Eastern Cooperative Oncology Group study.

PURPOSE: Fludarabine (2-fluoro-arabanoside-monophosphate) is a new antimetabolite chemotherapeutic agent. We performed a multicenter, phase II study of this drug in previously treated patients with refractory or relapsed non-Hodgkin's lymphoma (NHL) to determine its response rate by histologic classification. PATIENTS AND METHODS: Sixty-two assessable patients were given 18 mg/m2 by intravenous (IV) bolus injection daily for 5 days, every 28 days. Forty-eight percent had previously had one chemotherapy regimen, and the remainder had had two regimens; 42% had had radiation. RESULTS: Patients received 273 cycles of fludarabine chemotherapy, with a median of two cycles and ranging up to 25 cycles. Sixty patients were assessable for response, including nine complete responses (CRs; 15%) and nine partial responses (PRs; 15%). The response rate for patients with lower-grade histology was 52% (13 of 25); the greatest response rate was seen in those with follicular small cleaved-cell lymphoma, including seven of 11 treated. Five responders remain in unmaintained remission; the median survival of responders is greater than 30 months. Toxicity included mild neutropenia and a 10% incidence of grade 3 neurologic toxicity with occasional reversible visual and auditory changes. CONCLUSION: Fludarabine is active in patients with previously treated NHL (particularly low-grade histologies). Future studies will examine its activity in combination with other chemotherapeutic agents in previously untreated patients.

Report of the National Cancer Institute-sponsored workshop on definitions of diagnosis and response in acute myeloid leukemia.

The National Cancer Institute (NCI) sponsored a workshop to develop a set of standardized diagnostic and response criteria for acute myeloid leukemia (AML) clinical trials. The French-American-British (FAB) classification was retained for diagnosing AML, with the addition of patients with bone marrow morphologic features of a myelodysplastic syndrome and less than 30% bone marrow blasts, but with greater than or equal to 30% blasts in the peripheral blood. In this report, there are four important subgroups of AML not defined in the FAB classification that are discussed: undifferentiated acute leukemia, MO (AML lacking definitive myeloid differentiation by morphology or conventional cytochemistry but with ultrastructural or immunophenotypic evidence for AML), mixed lineage leukemia, and hypocellular AML. Definitions of response for clinical trials are presented to facilitate comparisons among different studies. Complete remission is considered the only response worth reporting in phase III trials, since lesser responses do not improve survival. Partial remissions may be of interest to identify active new agents in phase I and II studies. Monoclonal antibodies and cytogenetic studies are not part of the routine assessment of remission or reassessment at relapse, and their role in the evaluation of patients with AML is currently being evaluated in clinical trials. Although we recognize that some of the definitions in this report are arbitrary, generalized use of these guidelines will make results of clinical trials more comparable and interpretable.

Meta-analysis of chemotherapy versus combined modality treatment trials in Hodgkin's disease. International Database on Hodgkin's Disease Overview Study Group.

DESIGN: To perform a meta-analysis of all randomized trials that compared chemotherapy (CT) alone versus combined modality treatment (CT + radiotherapy [RT]) for which individual patient data could be made available. PATIENTS AND METHODS: Data on 1,740 patients treated on 14 different trials that included 16 relevant comparisons have been analysed. Eight comparisons were designed to evaluate the benefit of additional RT after the same CT (CT1 v CT1 + RT; additional RT design). Eight comparisons were designed to evaluate whether RT in a combined modality setting can be substituted by CT using either more cycles of the same CT or regimens that contain additional drugs (CT1 + CT2 v CT1 + RT or CT1 v CT2 + RT; parallel RT/CT design). RESULTS: Additional RT showed an 11% overall improvement in tumor control rate after 10 years (P = .0001; 95% confidence interval [CI], 4% to 18%). No difference could be detected with respect to overall survival (P = .57; 95% CI, -10% to 4%). In contrast, when combined modality treatment was compared with CT alone in the parallel-design trials, no difference could be detected in tumor control rates (P = .43; 95% CI, -6% to 9%), but overall survival was significantly better after 10 years in the group that did not receive RT (P = .045; 8% difference; 95% CI, 1% to 15%). There were significantly fewer fatal events among patients in continuous complete remission (relative risk [RR], 1.73; 95% CI, 1.17 to 2.53; P = .005) if no RT was given. CONCLUSION: Combined modality treatment in patients with advanced-stage Hodgkin's disease overall has a significantly inferior long-term survival outcome than CT alone if CT is given over an appropriate number of cycles. The role of RT in this setting is limited to specific indications.

High dose cytarabine: a review.

High dose cytarabine (HDARAC) therapy is used increasingly to treat hematologic malignancies. Recent data indicate that HDARAC at doses of 2-3 g/M2 every 12 hr x 10-12 doses is of comparable or greater efficacy in remission induction as standard doses of cytarabine in acute myelogenous leukemia. HDARAC can also produce remissions in individuals resistant to conventional doses. HDARAC-containing regimens are reported to result in substantially higher long-term, disease-free survival than previous approaches to post-remission therapy, but this has not yet been confirmed in controlled trials. HDARAC is also active in acute lymphocytic leukemia. Because intravenous HDARAC achieves high levels in the spinal fluid, it is useful to treat central nervous system leukemia and may provide adequate CNS prophylaxis in acute lymphocytic leukemia. HDARAC is reported to be active in advanced non-Hodgkin lymphomas and chronic myelogenous leukemia in acute phase; optimal use in these settings is under study. HDARAC has also been combined with other drugs. Randomized trials are needed to determine whether these combinations are more effective than HDARAC alone. Apart from potent myelosuppression, the dose-limiting toxicity of HDARAC is cerebellar damage. This occurs with increased frequency in patients greater than 50 years old. HDARAC is active in hematologic malignancies and may further improve therapeutic results if combined with other drugs.

Low dose cytosine arabinoside in myelodysplasia and acute myelogenous leukemia: a review.

We analyzed clinical trials of low dose cytosine arabinoside (LDARA-C) in 324 patients with acute myelogenous leukemia (AML) and 129 patients with myelodysplasia (MDS). Complete and partial remission rates were 31% and 18%, respectively, in patients with AML, and 24% and 27% in patients with MDS. Toxicity was primarily hematologic. Although in vitro data suggested that LDARA-C acts as a differentiating agent, clinical data generally indicate a cytotoxic mechanism. Given the lack of effective therapeutic options in MDS and high risk AML (patients greater than 65 years old, secondary AML), these data are encouraging. LDARA-C warrants further study, comparing continuous infusion with intermittent subcutaneous administration and comparing LDARA-C to conventional dose therapy.

Interleukin-2 therapy for advanced chronic myeloid leukemia.

Chronic myeloid leukemia (CML) is usually treated with either alpha-interferon or hydrea. Median survival is 6 years. Eventually, in most CML patients, the disease evolves to blast phase with clinical and morphologic characteristics of an acute leukemia. This phase is commonly associated with systemic symptoms and the appearance of new cytogenetic abnormalities. Therapy for this phase is of limited value, resulting in a mean survival of 4 months. We describe four consecutive patients seen at our clinic with advanced stage CML (three blast, one accelerated phase) who were treated with interleukin-2 (Proleukin). The mean survival in these patients was 22 months (range 9-35 months) and two are still alive 25 and 35 months after the start of therapy. One patient had a complete cytogenetic response and another a partial response. Toxicity was minimal and no patient had to discontinue therapy because of it.

Biologic heterogeneity in Philadelphia chromosome-positive acute leukemia with myeloid morphology: the Eastern Cooperative Oncology Group experience.

Evaluable karyotypes were available in 776 of 1148 adult patients who were entered on acute myeloid leukemia (AML) treatment protocols of the Eastern Cooperative Oncology Group. Among these, we found seven patients (0.9%) with t(9;22)(q34;q11), the Philadelphia (Ph) chromosome, in bone marrow metaphases. All fulfilled the FAB criteria for AML (three M0, two M1, two M2), although one patient presented with an additional, distinct lymphoid blast cell population. Chromosomal aberrations in addition to the Ph chromosome were seen in four patients (including two cases of monosomy 7). Molecular analysis by polymerase chain reaction in four patients tested revealed variable BCR/ABL transcript forms (ela2, b2a2, b3a2, b2a3+ela2). By immunophenotyping, all seven patients were myeloid based on the overall antigen expression pattern. However, all but one demonstrated lymphoid-associated antigens on the myeloid blast cells. The six evaluable patients failed to respond to treatment with a standard anthracycline/cytosine arabinoside-containing regimen. We conclude that the incidence of the Ph chromosome in AML is very low. Although both genotypically and phenotypically heterogenous, Ph chromosome-positive AML, represents a clinically distinct entity with poor outcome.

Phase II clinical trial of carboplatin in relapsed and refractory leukemia.

Carboplatin is a second-generation platinum complex drug which has demonstrated activity against a variety of neoplasms including acute leukemia, particularly when given by continuous intravenous (i.v.) infusion. Adults with acute myelogenous leukemia (AML) or acute lymphoblastic leukemia (ALL), either refractory or in first or second relapse, were given a continuous i.v. infusion of carboplatin at a dose of 315 mg/m2 daily for 5 days. A second course was given if the bone marrow at day 14 showed persistent leukemia. If the marrow was hypoplastic, treatment was delayed until marrow recovery was documented. Those with residual leukemia were given a second course. Those achieving complete remission (CR) were given an additional course as consolidation. Of the 46 eligible patients entered (36 AML and 10 ALL) eight achieved CR (17%) of which 6 were AML and 2 ALL. Of nine primary refractory patients, two achieved CR, one AML and one ALL. Excluding the inevaluable patients (protocol violations, patient refused further therapy, early deaths prior to day 14, the CR rate was eight of 28 (29%). All except two CRs required two courses of induction. The non-hematologic toxicity was minimal except for renal and auditory toxicity. Renal toxicity greater than grade 2 was seen in 17 patients and was associated with concomitant use of nephrotoxic antibiotics. In two patients, renal failure was a major factor in the cause of death. Ototoxicity was observed in 11 patients, but was grade 3 in only three. There were 18 deaths during the study. Fourteen died of infection, two died of infection and hemorrhage, one died of hemorrhage while aplastic, and one died of other causes. This trial indicates that carboplatin is an active agent in acute leukemia and warrants further investigation.