Effects of the epipodophyllotoxin VP-16-213 on cell cycle traverse, DNA synthesis, and DNA strand size in cultures of human leukemic lymphoblasts.

Serial studies of human leukemic lymphoblasts (CCRF-CEM line) cultured with 0.25 to 2.5 microM VP-16-213 for 0 to 6 hr indicated that the mechanism of cytotoxicity of this compound involves a primary effect on DNA. The most striking early change shown by flow cytometry in VP-16-213-treated cells was a delay in S-phase transit before arrest of cells in G2. Coinciding with this S-phase delay was a selective inhibition of thymidine incorporation into DNA as well as concentration-dependent scission of DNA strands. Using alkaline elution methods, we were able to detect DNA breakage at concentrations of VP-16-213 well below the level required to demonstrate kinetic effects or inhibition of DNA synthesis. These data suggest that DNA strand scission is the initial event in the sequence of kinetic and biosynthetic changes leading to growth inhibition and death of VP-16-213-treated cells. Inhibition of replicon initiation due to strand scission is a plausible explanation for the cytotoxic action of this podophyllotoxin derivative.

Clinical utility of initial terminal deoxynucleotidyl transferase determinations in childhood acute leukemias.

Terminal deoxynucleotidyl transferase (TDT) activity was measured in bone marrow lymphoblasts obtained at diagnosis from 168 consecutive patients with childhood acute leukemia. Absolute concentrations of TDT were increased (greater than or equal to 20 units/10(8) blasts) in samples from 98 of 112 assessable patients with acute lymphocyte leukemia (ALL). The values ranged from less than 1 to 1502 units/10(8) blasts with a median of 90 units contrasted with less than 1 to 219 units (median, 2.6 units) in studies of children without leukemia. Results of an immunofluorescence assay were in good agreement with enzymatic detection of the polymerase. Among 115 patients with adequate marrow smears, 105 had TDT-positive blasts. By contrast, in most children with acute myelogenous leukemia, TDT activity was either undetectable or less than 10 units/10(8) blasts. Although the highest levels of TDT were found in blasts with the common ALL phenotype, quantitative determinations were not significantly related to the major immunological subtypes of ALL or to morphological features or periodic acid-Schiff reactivity of the lymphoblasts. The probability that a newly diagnosed case of leukemia would be ALL was 90% if TDT levels were greater than 20 units/10(8) blasts. We conclude that absolute concentrations of TDT, as determined in this study, are of little value in identifying subclasses of ALL. The immunofluorescence assay, which is much less expensive and easier to perform than the enzyme assay, should prove useful for confirming the diagnosis of ALL and for detecting extramedullary sites of leukemic infiltration.

Failure of late intensification therapy to improve a poor result in childhood lymphoblastic leukemia.

This clinical study, begun in 1975, tested the efficacy of early and delayed intensification treatments in children with acute lymphoblastic leukemia. Regardless of presenting features, all patients received 4 weeks of conventional induction therapy with daily prednisone and weekly vincristine and daunorubicin. One-third were randomized to receive, in addition, two doses of asparaginase during induction therapy, while another one-third received four doses of both asparaginase and cytarabine after remission induction. Preventive central nervous system therapy uniformly included 2400 rads cranial irradiation and five doses of intrathecal methotrexate. Remissions were maintained with daily p.o. mercaptopurine and weekly i.v. methotrexate. Of the 277 assessable patients, 254 (92%) entered complete remission, and 102 (37%) remain clinically free of leukemia for 4.6 to 8.0 years (median, 6.3 years). The three treatment groups showed no significant differences in either remission induction rate or outcome, even when the analysis was based on risk assignment. A "late intensification" phase of therapy, added to the maintenance protocol for 65 patients who had been in continuous complete remission for 14 to 30 months, failed to extend remission durations, as judged from statistical comparison with matched controls (p = 0.84). When tested as a time-dependent covariate in the Cox proportional-hazards model, delayed intensification again showed no important effect on duration of complete remission. We conclude that limited early or aggressive late intensification of therapy, as described here, does not improve outcome in childhood acute lymphoblastic leukemia.

Acute nonlymphoblastic leukemia in infants: clinical presentation and outcome.

The presenting features and clinical outcome of acute nonlymphoblastic leukemia (ANLL) in infants and older children were compared to identify any differences that might suggest methods to improve therapy. Twelve of the 29 infants were boys and 17 were girls, with ages ranging from two days to 12 months (median, 7 months). By comparison with 222 patients greater than 1 year of age, infants were significantly more likely to have monoblastic or myelomonoblastic leukemia (P less than .0001), chloroma (P less than .0001), marked hepatomegaly (P = .001), and high leukocyte count (P = .005) and were less likely to have Auer rods (P less than .001). Each of these features except leukocyte count showed an association with infant ANLL in a multivariate analysis. Twenty-four (83%) of the infants attained a complete remission, a rate that was not significantly different from that of the older children. Even though infants had a significantly higher CNS relapse rate (P = .003), their event-free survival times were no different than those of older children (P = .74). Ten of the infants remain in initial complete remission for 5+ to 112+ months (median, 52+ months). Infants with ANLL did not have a poorer prognosis than older patients in our study; future protocols for this age group should emphasize more effective systemic therapy, preferably including an epipodophyllotoxin, as well as improved treatment for subclinical CNS leukemia.

Unexpectedly severe toxicity from intensive early treatment of childhood lymphoblastic leukemia.

In early 1984, we treated 13 consecutive patients with acute lymphoblastic leukemia (ALL) using an induction regimen of rapidly rotated combinations of prednisone, vincristine, asparaginase, teniposide (VM-26), cytosine arabinoside, and high-dose methotrexate (MTX) followed by leucovorin rescue. The intent of this clinical trial, designated Total Therapy Study XI, is to test the hypothesis that greater initial leukemia cell kill will decrease opportunities for the development of drug-resistant mutants, with resultant improvement in the length of disease-free survival. Five patients experienced life-threatening gastrointestinal toxicity within three weeks of the start of treatment. One died. Three other patients had severe abdominal pain, abdominal distention, diarrhea, and weight loss, but not gastrointestinal bleeding. In the remaining five patients, toxicity was rapidly reversible, and each child was able to complete the planned course of chemotherapy. The study was then amended to switch high-dose MTX from the induction phase to the consolidation phase, allowing at least one week for mucosal recovery. Among the next 28 patients who were treated, none showed evidence of severe gastrointestinal toxicity. Patients now receive high-dose MTX alone as consolidation therapy and are tolerating it adequately. Drug timing should be examined critically when intensified multiple-agent regimens are being devised for initial treatment of ALL.

Clinical relevance of lymphoblast biological features in children with acute lymphoblastic leukemia.

Improvements in therapy for childhood acute lymphoblastic leukemia (ALL) have led us to reevaluate the prognostic significance of lymphoblast characteristics at diagnosis. From application of univariate and multivariate statistical methods, we determined the relationship of five blast cell features to treatment outcome in 250 patients who were enrolled in two clinical trials at this center from May 1979 through April 1982. Karyotype ploidy, lymphoblast morphology, and immunophenotype were each significantly related to prognosis as measured by time to failure, while periodic acid-Schiff reactivity and glucocorticoid receptor number lacked prognostic implication for this patient population. In addition, clinical features of initial WBC count, age, and race were also significant independent variables in predicting treatment response. By multivariate analysis, both ploidy and morphology contributed prognostic information to a clinical model based on WBC count, age, and race. If the model was adjusted for impact of ploidy, however, French-American-British morphology no longer contributed additional prognostic information. Our findings suggest that many traditional biological features used to estimate prognosis in ALL can be discarded in favor of clinical features (leukocyte count, age, and race) and cytogenetics (ploidy) for planning of future clinical trials.

Prognostic importance of cytogenetic subgroups in de novo pediatric acute nonlymphocytic leukemia.

Reports of close associations between recurring chromosomal abnormalities and the clinical behavior of acute nonlymphocytic leukemia (ANLL) have stimulated efforts to define this disease in cytogenetic terms. Here we report on the leukemic cell karyotypes of 155 children with ANLL who were treated from 1980 to 1987 in consecutive programs of chemotherapy at this institution. Of 121 cases with adequate banding, 20% were normal, 30% had miscellaneous clonal abnormalities, and 50% were classified into known cytogenetic subgroups: inv(16)/del(16q) (n = 15), t(8; 21) (n = 14), t(15;17) (n = 9), t(9;11) (n = 9), t(11;V)/del(11q) (n = 7) and -7/del(7q) (n = 6). The inv(16)/del(16q) cases showed a nearly equal distribution of myelocytic and monocytic French-American-British (FAB) subtypes; only four of these patients presented with M4Eo morphology. Despite a 100% remission induction rate, patients with inv(16)/del(16q)-positive ANLL fared no better overall than the entire group; only 40% of this subgroup were event-free survivors at 2 years from diagnosis (P = .23). Patients with inv(16)/del(16q) frequently had CNS involvement at diagnosis (eight of 15) or initially relapsed in this site (three of eight). Event-free survival (EFS) was clearly superior for young patients with FAB M5 leukemia and the t(9;11) (P = .041). These patients were clinically indistinguishable from others with the FAB disease subtype, yet their responses to etoposide-containing therapies were noteworthy. By contrast, children with structural abnormalities involving 11q23, other than t(9;11), were infants (median age, 6 months) with FAB M4 or M5 leukemia, hyperleukocytosis, and frequent coagulation abnormalities. Patients with such changes [t(11;V) or del(11q)] relapsed early during postremission therapy: none remained disease-free more than 16 months from diagnosis. Because of resistant leukemia, patients with monosomy 7/del(7q) had a poor remission induction rate (17%; P = .0015); patients with the t(15;17) were also poor responders to induction therapy (44%; P = 0.02) because of hemorrhagic deaths. These results identify several cytogenetic subtypes of pediatric ANLL that may represent unique disease processes for which more effective early cytoreduction [-7/del(7q), t(11;V)], better supportive care measures [t(15;17)], or more effective CNS prophylaxis [inv(16)/del(16q)] would be warranted.

Allogeneic bone marrow transplantation in a program of intensive sequential chemotherapy for children and young adults with acute nonlymphocytic leukemia in first remission.

Eighty-seven consecutive children and young adults with acute nonlymphocytic leukemia (ANLL) were treated uniformly with induction chemotherapy based on daunorubicin and cytarabine (ara-C), with the addition of etoposide (VP-16) and azacytidine (5-Az) for refractory patients. Of the 65 patients who entered complete remission, 42 were eligible for assessment of response to intensive chemotherapy consisting of four pairs of drugs administered in sequential fashion. Nineteen others with available histocompatibility locus antigen (HLA)-compatible donors were assigned to receive allogeneic bone marrow transplants within 16 weeks from their dates of complete remission. Durations of continuous complete remission (CCR) in the two groups were not significantly different at a median follow-up time of 6 years (P = .30 by log-rank analysis). Kaplan-Meier estimates of CCR probabilities (+/- SE) at 6 years were 43% +/- 13% (transplantation) and 31% +/- 7% (sequential chemotherapy). Postremission failures in the sequential chemotherapy group resulted from bone marrow relapse in 23 of 29 patients (79%), whereas in the transplantation group, failures were equally divided between marrow relapse and transplantation-related complications of graft-versus-host disease (GVHD) or infection due to the immunosuppressive effects of ablative chemotherapy. Comparison of hematologic remission curves indicated a significant advantage for marrow transplantation in terms of systemic leukemia control (P = .06). Thus, in programs of intensive chemotherapy of the type described here, allogeneic marrow transplantation should be seriously considered as alternative therapy for patients in first remission who have an HLA-matched sibling donor, provided that effective methods for control of transplant-related complications are available.

Teniposide plus cytarabine improves outcome in childhood acute lymphoblastic leukemia presenting with a leukocyte count greater than or equal to 100 x 10(9)/L.

Childhood acute lymphoblastic leukemia with an initial leukocyte count greater than or equal to 100 X 10(9)/L responds poorly to conventional chemotherapy. To extend event-free survival (EFS) in this disease, we devised a protocol that specifies intensive 2-week courses of teniposide (VM-26, 165 mg/m2) plus cytarabine (ara-C, 300 mg/m2), before and immediately after standard 4-week remission induction therapy with prednisone, vincristine, and L-asparaginase. The VM-26 and ara-C combination was also administered intermittently for the first year of continuation treatment with oral 6-mercaptopurine and methotrexate. CNS prophylaxis consisted of periodic intrathecal (IT) injections of methotrexate and delayed cranial irradiation. At a median follow-up of 4 years, the estimated EFS rate for 57 consecutive patients with leukocyte counts of 100 to 1,000 X 10(9)/L was 44%, compared with 10% for matched controls (P less than .001). Remission induction rates in the two groups were similar (82% v 72%, P = .16). Twenty-five patients in the VM-26/ara-C group have survived without adverse events for 2.7 to 6.8 years, whereas only nine of the controls achieved more than a year of EFS. The most common complications during early treatment were acute hyperkalemia from rapid tumor cell lysis and infections due to prolonged marrow aplasia. Continuation chemotherapy was well tolerated. We conclude that VM-26 plus ara-C, added to each phase of an otherwise basic regimen of chemotherapy, will substantially improve prognosis in this high-risk form of childhood leukemia.

Correlation of karyotype and immunophenotype in childhood acute lymphoblastic leukemia.

To correlate leukemic cell karyotype with immunophenotype, we studied 364 children with acute lymphoblastic leukemia (ALL). A prognostically favorable cytogenetic feature, hyperdiploidy greater than 50 chromosomes, was found in 33% of cases classified as common ALL antigen positive (CALLA+) early pre-B (common) ALL, in contrast to 18% of pre-B cases (P = .012), 5% of T cell cases (P less than .001), and none of the B cell cases (P less than .001) or cases of CALLA negative (CALLA-) early pre-B ALL (P = .002). The frequency of translocations, an adverse cytogenetic feature, was significantly lower in CALLA+ early pre-B ALL cases (35%) than in B cell (100%; P less than .0001), pre-B (59%; P less than .001), or CALLA- early pre-B (62%; P = .016) cases. Thus, patterns of chromosomal change differ widely among the major immunophenotypic groups of ALL and may account for reported differences in responsiveness to treatment.

The t(10;11)(p14;q21) translocation in three children with acute myeloblastic leukemia.

A total of 161 cases of pediatric de novo acute myeloblastic leukemia (AML) have been reviewed, for which complete karyotyping was available and three cases (2%) were identified with t(10;11)(p14;q21). Two of the three children were infants with monoblastic (FAB M5) leukemia and the third was an adolescent with undifferentiated myeloid (FAB M1) leukemia. Both infants presented with increased levels of lactate dehydrogenase. None of these cases had increased eosinophils. One of the infants is in remission 18+ months after diagnosis and intensive chemotherapy; the two other children attained brief initial remissions but succumbed to their disease within 11 months of diagnosis. The prognosis of such children appears to be similar to that of cases of AML lacking this translocation.

Childhood acute promyelocytic leukemia: a rare variant of nonlymphoid leukemia with distinctive clinical and biologic features.

Of 251 consecutive cases of childhood acute nonlymphocytic leukemia (ANLL) seen at St. Jude Children's Research Hospital over a 12-year period, 16 (6.4%) were classified as promyelocytic according to the French-American-British definition. Patients with this form of leukemia were older at diagnosis than the group representing all other ANLL subtypes (median age, 14.8 vs. 9.0 years); they had lower leukocyte counts (median, 4.5 vs. 25.9 x 10(9)/liter), and a higher percentage were girls (68% vs. 44%). They also were much more likely to have a coagulation abnormality (75% vs. 13%). Only 44% of the promyelocytic group achieved complete remission, compared with 79% of the remaining patients (p = 0.001); however, after a median follow-up of 3.5 years, all but two of the responding patients with promyelocytic leukemia remain in complete remission. The majority of induction failures in the promyelocytic group (six of nine) resulted from complications that developed during periods of marrow hypoplasia or before hypoplasia was induced; whereas in the comparison group, more than half of the patients who failed had evidence of absolute or relative drug resistance. It is concluded that acute promyelocytic leukemia in children differs sufficiently from other subtypes of childhood ANLL to justify clinical trials of selective therapy. Recommendations for the use of heparin and blood component support in these patients are given.

Impact of treatment efficacy on the prognostic value of glucocorticoid receptor levels in childhood acute lymphoblastic leukemia.

Glucocorticoid receptor (GR) levels were quantitated in leukemic blasts from bone marrow aspirates of 249 children with acute lymphoblastic leukemia (ALL) who were entered on two St. Jude Total Therapy Studies. Of these, 235 were evaluable for analysis of the relation of GR levels to clinical outcome. For the 42 patients in the earlier Total Therapy Study IX, lower GR levels (less than 16,000 sites/cell) were associated with both induction failure and more frequent relapse (p less than 0.01) [Cancer Research, Vol. 42, p. 4801 (1982)]. When patients with 'high-risk' features (leukocyte count greater than 100 X 10(3)/mm3, positive erythrocyte rosette test, central nervous system involvement, and mediastinal mass) were excluded, lower receptor levels were still associated with early and more frequent relapse (p less than 0.02). The other 193 evaluable patients were consecutively admitted to Total Therapy Study X, in which patients with 'standard-risk' or 'high-risk' features were assigned to separate protocols--XS and XH, respectively. Induction chemotherapy in both protocols consisted of prednisone, vincristine and L-asparaginase; patients in the XH protocol received additional epipodophyllotoxin (VM-26) and cytosine arabinoside twice a week for 2 weeks preceding the conventional induction therapy. To compare the prognostic value of GR level in Study X with that of Study IX (which included both 'high-risk' and 'standard-risk' patients but did not separate them into different protocol groups), children in the XH and XS protocols were analysed together. The proportion of patients with 'standard-risk' features was the same in the two studies: 69% in Study IX and 73% in Study X. In Study X, which had a significantly better treatment result (p less than 0.001), lower receptor levels were not associated with induction failure, but were correlated with more frequent relapse (p less than 0.05). When patients in XH and XS protocols were analysed separately, however, receptor levels were no longer related to treatment outcome. Thus, GR level in childhood ALL has prognostic value, but it is not an independent factor and its importance is related to the efficacy of treatment.

Glucocorticoid receptors in childhood acute non-lymphocytic leukemia.

Using a whole-cell assay, we found glucocorticoid receptors (GR) in all 43 consecutive assessable children with newly diagnosed acute non-lymphocytic leukemia (ANLL). The receptor levels ranged from 2146 to 81,308 sites/cell (median = 18,105); these results were similar to those for acute lymphocytic leukemia. Receptor levels were not related to any of these clinical or biological features at diagnosis: age, sex, race, initial leukocyte count, liver or spleen size, presence of CNS disease or Auer rods, [3H] thymidine ( [3H]TdR) labelling index, French-American-British morphology or terminal deoxynucleotidyl transferase activity. Receptor levels also were not related to the initial treatment response or remission duration after therapy that did not include a glucocorticoid. We conclude that GR level has no clinical utility as a marker protein in childhood ANLL.

Comparison of two clonogenic assays for determining cell growth patterns in childhood acute myeloid leukemia: relationships to clinical findings and prognosis.

Bone marrow cells from 99 patients with acute myeloid leukemia were cloned in either agar stimulated by leukocyte feeder layers (AG/F) or methylcellulose supplemented with medium conditioned by phytohemagglutinin stimulation of leukocytes (MC/P). Although cell growth in the two systems was correlated (r = 0.74, p less than 0.0001), there was increased formation and size of clusters and colonies in AG/F, suggesting that the clonogenic cells from children with AML are more readily assayed in AG/F. The number and size of clones in either system did not show a relationship to the morphologic subtype of leukemia. Depending on the scoring system used, increased growth in MC/P was related to abnormal karyotype. Also dependent on scoring system, the ability of leukemic cells to form small clusters in AG/F was associated with resistance to induction therapy: cells of patients with resistant disease were more likely to produce small clusters (p = 0.02). Our results suggest that clonogenic cells from children with AML grow more readily in AG/F than in MC/P, but that neither culture system supports the growth of cells from all patients. Depending on scoring criteria, in-vitro growth patterns in AG/F correlate with response to induction therapy.

Unfavorable presenting clinical and laboratory features are associated with CALLA-negative non-T, non-B lymphoblastic leukemia in children.

Twenty-four (5.7%) of 424 children with newly diagnosed acute lymphoblastic leukemia (ALL) were found to have blast cells that expressed HLA-DR antigens but not the common ALL antigen (CALLA), E-rosette receptors, T-cell antigens, or cytoplasmic or surface immunoglobulins. Each of the eight cases tested expressed the B-cell associated antigen B4, but not B1 or B2 antigen. Myeloid-associated antigens were not present in any of the 10 cases tested. By comparison with common (CALLA+ B-cell precursor) ALL, patients having this immunophenotype were more likely to be children less than 2 yr of age (p less than 0.001), to have higher initial leukocyte counts (p less than 0.001), and to have blast cells with a DNA index less than 1.16 (p = 0.05), a pseudodiploid karyotype (p = 0.01) and a chromosomal translocation (p = 0.003). The presence of any chromosomal translocation in these CALLA- ALL was related to measures of increased leukemic cell burden including higher leukocyte counts, larger liver and spleen sizes and higher serum lactic dehydrogenase levels. While the patients were entered into several treatment arms of two protocols, the CALLA- cases appeared to have lower remission rate (p = 0.06) and shorter event-free survival time (p = 0.05) than did those with common ALL. The association with clinical and laboratory features of known adverse prognostic significance provides some explanation for the poor treatment outcome of CALLA- ALL.

Monitoring serum aminoglycoside concentrations in children with amphotericin B nephrotoxicity.

We prospectively studied the effect of amphotericin B therapy on aminoglycoside clearance in 20 consecutive children during the remission-induction phase of chemotherapy for acute myelocytic leukemia. Increases (greater than 50%) in the half-life for aminoglycoside excretion were not associated with antileukemic or aminoglycoside therapy alone but occurred in 12 of 17 children when amphotericin B was added to the antimicrobial regimen. Seven children had impaired aminoglycoside clearance without increases (greater than 50%) in serum creatinine; hence the resulting adjustments in aminoglycoside dosage would not have been made had we relied solely on serial measurements of serum creatinine. Evidence for increased excretion of the renal enzymes N-acetyl-beta-D-glucosaminidase and alanine aminopeptidase during amphotericin B therapy suggested that damage to proximal tubular cells may contribute to the renal impairment that has been associated with this drug. Our findings underscore the value of monitoring serum aminoglycoside concentrations in children being treated with amphotericin B.

Periodic acid-Schiff stain in childhood acute lymphoblastic leukemia: lack of independent prognostic value.

The authors reviewed 250 consecutive children with ALL to determine if the periodic acid-Schiff (PAS) score was a useful, independent predictor of time to failure. PAS stains were scored from 0 to 400 and divided into low- and high-score groups using a variance-ratio test (F test) to optimize any effect of PAS on prognosis. Although the effect of PAS score considered alone approached significance for time to failure, the PAS score lost all significance when the patients were divided into standard-risk and high-risk groups on the basis of peripheral white count, central nervous system involvement, mediastinal mass, or E-rosette positivity at diagnosis. A Cox regression analysis was performed on a subgroup of 198 patients for whom cytogenetic studies were also available. The PAS score again approached the level of significance when considered alone but was of no significance after the effects of peripheral white count, pseudodiploidy, mediastinal mass, and E-rosette positivity were removed. The authors conclude that the PAS stain has no independent prognostic significance in childhood ALL.

Auer rods in mature granulocytes: a unique morphologic feature of acute myelogenous leukemia with maturation.

The authors observed Auer rods in mature neutrophils, bands, and metamyelocytes in two children with AML with maturation (FAB M2). In order to determine if this rare finding was characteristic of AML M2, the authors reviewed the bone marrows and peripheral blood smears of 50 children with AML M2 and 50 children with other AML subtypes (FAB M1, M3, M4, and M5). They found Auer rods in mature neutrophils and band forms in 10 of 50 patients (20%) with AML M2 but in none of 50 patients with non-M2 AML. This finding was not related to the frequency of mature cells in the specimen or to the frequency of Auer rods in blasts. Cytogenetics did not show a consistent abnormality. The presence of Auer rods in mature granulocytes is unique to FAB M2 AML in the authors' series and supports the concept that in AML at least some of the mature myeloid cells are involved in the leukemic process.

Cytogenetics of childhood acute nonlymphocytic leukemia.

Interest in more precise subclassification of the acute leukemias by cytogenetic criteria led us to identify and characterize the full range of chromosomal abnormalities in 121 children with de novo acute nonlymphocytic leukemia (ANLL). Only 21% of the cases had normal karyotypes; 62% had consistent or recurrent alterations, most commonly inv(16) or del(16), t(8;21), t(15;17), t(9;11), t(11;V) or del(11), and -7 or 7q-; and 17% had miscellaneous, apparently random, clonal abnormalities. Statistically significant associations between chromosomal abnormalities and the morphologic/cytochemical subtypes of ANLL, defined by criteria of the French-American-British (FAB) cooperative group were demonstrated for the t(8;21) in M1 and M2 leukemia, t(15;17) in M3, t(9;11) in M5, and translocations involving 11q23 other than t(9;11) [t(11;V)] or del(11q) in M4 and M5. The chromosome 16 inversion was not restricted to the M4 subtype, as is generally reported, and was not uniformly associated with increased and/or abnormal marrow eosinophils. None of these 121 cases were characterized by the Philadelphia chromosome, nor did any have the t(6;9), t(16;16), or inv(3), which have been noted previously in this disease. In addition to confirming several recognized correlations between recurrent structural chromosome abnormalities and FAB subtypes, this study identified novel abnormalities that have not been reported by others. It also disclosed an unusual heterogeneity of chromosome 16 abnormalities with respect to their distribution among FAB subtypes, their association with marrow eosinophilia, and their participation with other chromosomes in translocations.