Different target range and cytotoxic specificity of adaphostin and 17-allylamino-17-demethoxygeldanamycin in imatinib-resistant and sensitive cell lines.

Imatinib mesylate is a selective inhibitor of the oncogenic tyrosine kinase, Bcr-Abl, and is widely used as a first-line treatment for chronic myeloid leukaemia (CML). Prolonged monotherapy is frequently associated with patients becoming refractory to imatinib. Therefore, there is considerable interest in small molecule inhibitors which may be used either as replacements or as adjuncts to existing imatinib therapy. For this purpose, it is most likely that drugs which do not share imatinib's mechanism of action will be most valuable. We compared two such compounds with different modes of action, adaphostin and 17-allylamino-17-demethoxygeldanamycin (17-AAG), for their cytotoxic effect and ability to induce the downregulation of cellular proteins in a murine haemopoietic cell line transformed with human p210(Bcr-Abl), and two subclones resistant to imatinib owing to an Abl-kinase domain mutation (E255K) or amplification of the BCR-ABL gene, respectively. We found that, whereas 17-AAG selectively killed Bcr-Abl-positive cells and inhibited proteins dependent on heat-shock protein 90 for their stability (p210(Bcr-Abl) and Akt), adaphostin induced the downregulation of multiple cell-signalling proteins (p210(Bcr-Abl), Akt, Bcr, Abl and STAT5a) and was cytotoxic to both Bcr-Abl-positive and -negative cells. We suggest that both compounds may prove useful in the treatment of CML but caution that undesirable side-effects may result from the inhibition of multiple cell signalling proteins.

Overexpression of the heat-shock protein 70 is associated to imatinib resistance in chronic myeloid leukemia.

Imatinib is an effective therapy for chronic myeloid leukemia (CML), a myeloproliferative disorder characterized by the expression of the recombinant oncoprotein Bcr-Abl. In this investigation, we studied an imatinib-resistant cell line (K562-r) generated from the K562 cell line in which none of the previously described mechanisms of resistance had been detected. A threefold increase in the expression of the heat-shock protein 70 (Hsp70) was detected in these cells. This increase was not associated to heat-shock transcription factor-1 (HSF-1) overexpression or activation. RNA silencing of Hsp70 decreased dramatically its expression (90%), and was accompanied by a 34% reduction in cell viability. Overexpression of Hsp70 in the imatinib-sensitive K562 line induced resistance to imatinib as detected by a large reduction in cell death in the presence of 1 muM of imatinib. Hsp70 level was also increased in blast cells of CML patients resistant to imatinib, whereas the level remained low in responding patients. Taken together, the results demonstrate that overexpression of Hsp70 can lead to both in vitro and in vivo resistance to imatinib in CML cells. Moreover, the overexpression of Hsp70 detected in imatinib-resistant CML patients supports this mechanism and identifies potentially a marker and a therapeutic target of CML evolution.

BCR-ABL activity and its response to drugs can be determined in CD34+ CML stem cells by CrkL phosphorylation status using flow cytometry.

In chronic myeloid leukaemia, CD34(+) stem/progenitor cells appear resistant to imatinib mesylate (IM) in vitro and in vivo. To investigate the underlying mechanism(s) of IM resistance, it is essential to quantify Bcr-Abl kinase status at the stem cell level. We developed a flow cytometry method to measure CrkL phosphorylation (P-CrkL) in samples with <10(4) cells. The method was first validated in wild-type (K562) and mutant (BAF3) BCR-ABL(+) as well as BCR-ABL(-) (HL60) cell lines. In response to increasing IM concentration, there was a linear reduction in P-CrkL, which was Bcr-Abl specific and correlated with known resistance. The results were comparable to those from Western blotting. The method also proved to be reproducible with small samples of normal and Ph(+) CD34(+) cells and was able to discriminate between Ph(-), sensitive and resistant Ph(+) cells. This assay should now enable investigators to unravel the mechanism(s) of IM resistance in stem cells.

The presence of a BCR-ABL mutant allele in CML does not always explain clinical resistance to imatinib.

The expansion of a leukemia clone bearing a Bcr-Abl kinase domain mutation is associated with acquired resistance to imatinib and may also predict disease progression in patients with Philadelphia-positive chronic myeloid leukemia (CML). Here we report results of pyrosequencing to quantitate the non-mutated and mutant alleles in 12 CML patients monitored over periods ranging from 11 to 58 months, and describe three contrasting kinetic patterns: Group 1 - in four patients total BCR-ABL transcript numbers remained high with the mutant allele predominating; Group 2 - in four patients the total number of BCR-ABL transcripts fell to low levels but the mutant allele predominated; and Group 3 - in four other patients the total level of transcripts remained high (n = 2) or fell (n = 2) but the mutant clone persisted at relatively low level. In Group 2 the mutant leukemia clone was presumably still relatively sensitive to imatinib but in Group 1 the leukemia could be classified as resistant. In Group 3 patients the imatinib sensitivity of the leukemia was variable. We conclude that a mutant clone does not necessarily have a proliferative advantage and its presence does not always account for resistance to imatinib. Other mechanisms underlie resistance in at least some patients.

Zoledronate inhibits proliferation and induces apoptosis of imatinib-resistant chronic myeloid leukaemia cells.

Although imatinib mesylate has revolutionized the treatment of chronic myeloid leukaemia (CML), resistance to the drug, manifesting as relapse after an initial response or persistence of disease, remains a therapeutic challenge. In order to overcome this, alternative or additional targeting of signaling pathways downstream of Bcr-Abl may provide the best option for improving clinical response. Bisphosphonates, such as zoledronate, have been shown to inhibit the oncogenicity of Ras, an important downstream effector of Bcr-Abl. In this study, we show that zoledronate is equally effective in inhibiting the proliferation and clonogenicity of both imatinib-sensitive and -resistant CML cells, regardless of their mechanism of resistance. This is achieved by the induction of S-phase cell cycle arrest and apoptosis, through the inhibition of prenylation of Ras and Ras-related proteins by zoledronate. The combination of imatinib and zoledronate also augmented the activity of either drug alone and this occurred in imatinib-resistant CML cells as well. Since zoledronate is already available for clinical use, these results suggest that it may be an effective addition to the armamentarium of drugs for the treatment of CML.

Synergistic activity of imatinib and 17-AAG in imatinib-resistant CML cells overexpressing BCR-ABL--Inhibition of P-glycoprotein function by 17-AAG.

Overexpression of BCR-ABL and P-glycoprotein (Pgp) are two of the known mechanisms of imatinib resistance. As combination therapy may allow to overcome drug resistance, we investigated the effect of combination treatment with imatinib and 17-allylamino-17-demethoxygeldanamycin (17-AAG), a heat-shock protein 90 (Hsp90) inhibitor, on different imatinib-sensitive and imatinib-resistant CML cell lines. In imatinib-sensitive cells, combination index (CI) values obtained using the method of Chou and Talalay indicated additive (CI=1) or marginally antagonistic (CI>1) effects following simultaneous treatment with imatinib and 17-AAG. In imatinib-resistant cells both drugs acted synergistically (CI<1). In primary chronic-phase CML cells additive or synergistic effects of the combination of imatinib plus 17-AAG were discernible. Annexin V/propidium iodide staining showed that the activity of imatinib plus 17-AAG is mediated by apoptosis. Combination treatment with imatinib plus 17-AAG was more effective in reducing the BCR-ABL protein level than 17-AAG alone. Monotherapy with 17-AAG decreased P-glycoprotein activity, which may increase intracellular imatinib levels and contribute to the sensitization of CML cells to imatinib. The results suggest that combination of imatinib and 17-AAG may be useful to overcome imatinib resistance in a clinical setting.

BCR-ABL tyrosine kinase activity regulates the expression of multiple genes implicated in the pathogenesis of chronic myeloid leukemia.

The BCR-ABL chimeric protein is thought to play a central role in the pathogenesis of Philadelphia (Ph) chromosome-positive leukemias, notably chronic myeloid leukemia (CML). There is compelling evidence that malignant transformation by BCR-ABL is critically dependent on its protein tyrosine kinase (PTK) activity. As a result, multiple signaling pathways are activated in a kinase-dependent manner, and thus the activation of such pathways may affect the expression of genes that confer the malignant phenotype. In this study, we used differential display to investigate the alterations of gene expression in BV173, a CML cell line derived from lymphoid blast crisis, after exposure to ST1571, which selectively inhibits ABL PTK activity. We show that the expression of a set of 12 genes is correlated with the kinase activity and that the profile of these genes reflects mechanisms implicated in the pathogenesis of CML. Several of the genes show a consistent pattern of altered regulation in all Ph-positive lymphoid cell lines, whereas others appear to be unique to BV173 cells. We conclude that BCR-ABL PTK activity drives the expression of specific target genes that contribute to the malignant transformation of Ph-positive cells. The identification of downstream molecules with a consistent regulation pattern may provide suitable targets for therapeutic intervention in the future.

Selective induction of leukemia-associated fusion genes by high-dose ionizing radiation.

There is strong clinical and epidemiological evidence that ionizing radiation can cause leukemia by inducing DNA damage. This crucial initiation event is believed to be the result of random DNA breakage and misrepair, whereas the subsequent steps, promotion and progression, must rely on mechanisms of selective pressure to provide the expanding leukemic population with its proliferative/renewal advantage. To investigate the susceptibility of human cells to external agents at the genetic recombination stage of leukemogenesis, we subjected two hematopoietic cell lines, KG1 and HL60, to high doses of gamma-irradiation. The irradiation induced the formation of fusion genes characteristic of leukemia in both cell lines, but at a much higher frequency in KG1 than in HL60. In KG1 cells, the AML1-ETO hybrid gene [associated with the t(8;21) translocation of acute myeloid leukemia] occurred significantly more often than the BCR-ABL [associated with t(9;22) chronic myeloid leukemia] or the DEK-CAN [associated with t(6;9) acute myeloid leukemia] fusion genes. These findings support the notion that ionizing radiation can directly generate leukemia-specific fusion genes but emphasize the differing susceptibility of different cell populations and the differing frequency with which the various fusion genes are formed. The selectivity observed at the primary level of gene fusion formation may explain at least in part the differential risk for development of some but not other forms of leukemia after high-dose radiation exposure.

Direct relation between BCR-ABL tyrosine kinase activity and cyclin D2 expression in lymphoblasts.

Leukemia cells bearing the Philadelphia (Ph) chromosome express a Bcr-Abl fusion protein with deregulated protein tyrosine kinase (PTK) activity, which plays a central role in the malignant transformation. Many different signal transduction pathways are activated by Bcr-Abl, but little is known about their downstream targets in specific cell lineages. We show here that Ph-positive cell lines as well as primary cells derived from chronic myeloid leukemia (CML) in lymphoid blast crisis or from acute lymphoblastic leukemia (ALL) consistently express high levels of cyclin D2, whereas expression of this protein is low or absent in comparable Ph-negative lines and Ph-positive myeloid lines. Inhibition of Bcr-Abl with STI571 resulted in down-regulation of cyclin D2 and reduction of the number of cells in S phase, although complete G1 arrest was not induced. The expression of cyclin D2 in Ph-positive lymphoblasts was mediated via the phosphatidyl-inositol-3 kinase pathway. Analogous results were seen in murine BaF/3 cells transfected with a BCR-ABL expression vector. In contrast to the human cell lines, murine Baf/BCR-ABL cells exposed to STI571 inhibitor were all arrested in G1. This arrest could be abrogated by exogenous expression of cyclin D2 from a transfected cDNA construct. We conclude that a direct connection exists between Bcr-Abl PTK activity and cell cycle progression in which cyclin D2 plays a critical role. However, cell cycle progression in human Ph-positive lymphoid cells is not entirely dependent on Bcr-Abl PTK, and additional genetic lesions must be present.

The molecular biology of chronic myeloid leukaemia.

Chronic myeloid leukaemia (CML) is characterized cytogenetically by a t(9;22)(q34;ql1) reciprocal translocation which gives origin to a hybrid BCR-ABL gene, encoding a p2lO(BCR-ABL) fusion protein with elevated tyrosine kinase activity and transforming abilities. The t(9;22) was suggested to be associated with genomic imprinting of centromeric regions of chromosomes 9 and 22, but the genes directly affected by the translocation, ABL and BCR, were shown not to be imprinted. For most diagnostic and research purposes the BCR-ABL gene can be efficiently identified by reverse-transcription and polymerase chain reaction (RT/PCR) amplification of its fusion transcripts, which can be quantified by competitive PCR and similar assays for assessment of residual disease in the follow-up of therapy. In the great majority of CML patients the BCR-ABL transcripts exhibit a b2a2 and/or a b3a2 junction; in rare cases, the only detectable BCR-ABL transcripts have unusual junctions, such as b2a3, b3a3, e1a2 or e6a2. There is a recent suggestion that the BCR-ABL gene may not be always 'functional', since extremely low levels of BCR-ABL transcripts can be found in leucocytes from normal individuals and, conversely, it appears that no BCR-ABL transcription can be detected in a proportion of Ph-positive haematopoietic progenitors from some CML patients. The role, if any, of the reciprocal ABL-BCR hybrid gene in CML is unknown. Although its mRNA message is in frame, no ABL-BCR fusion protein has yet been identified in CML patients. The blast crisis of CML has been variably associated with abnormalities of proto-oncogenes, such as RAS and MYC, or of tumour suppressor genes, in particular RB, p53 and p16, or with the generation of chimeric transcription factors, as in the AML1-EVI1 gene fusion. It is likely, therefore, that multiple and alternative molecular defects, as opposed to a single universal mechanism, underlie the acute transformation of the disease.

Specific point mutations that activate v-abl are not found in Philadelphia-negative chronic myeloid leukaemia, Philadelphia-negative acute lymphoblastic leukaemia or blast transformation of chronic myeloid leukaemia.

The involvement of the BCRlABL fusion gene in patients with Philadelphia (Ph) chromosome positive chronic myeloid leukaemia (CML) and acute lymphoblastic leukaemia (ALL) is well characterised, but the molecular events underlying the cases of Ph-negative CML and ALL that lack BCR gene involvement and those that cause transformation of Ph-positive CML are unknown. The murine ABL gene can be activated by genetic events that do not involve the BCR gene, including the introduction of two specific point mutations in exons VII and XI respectively, as found in the homologous sequence of the v-abl oncogene. We therefore sought evidence for analogous point mutations in the ABL gene in patients with Ph-negative, BCR-negative CML (n = 25), Ph-negative ALL (n = 18) and in Ph-positive CML in transformation (n = 28). We used restriction fragment length polymorphism and single strand conformational polymorphism techniques to analyse DNA amplified fragments of selected ABL coding regions from leukaemia cells. We identified only normal wild-type DNA sequences. The absence of these transforming point mutations does not exclude the possibility that the ABL gene in such patients could be activated by other means.

Balanced parental contribution to the ABL component of the BCR-ABL gene in chronic myeloid leukemia.

Genomic imprinting has recently been associated with the reciprocal t(9;22) chromosome translocation of chronic myeloid leukaemia (CML). This translocation gives rise to a 22q-, or Philadelphia (Ph), chromosome and a derivative 9q+. Based on heterochromatin polymorphisms, it was reported that the former is of maternal and the latter of paternal origin in every case of CML. This parental bias led to the hypothesis that the genes disrupted by the translocation, BCR and ABL, were themselves imprinted, and that in CML the BCR-ABL gene was formed by BCR sequences of maternal and ABL sequences of paternal origin. We have identified a BstNl restriction fragment length polymorphism in the ABL coding sequence which enabled us to investigate directly the expression and inheritance of the two ABL alleles in heterozygous CML patients. Amplification of the specific BCR-ABL and normal ABL mRNA messages by reverse transcriptase-polymerase chain reaction in these patients showed that the ABL moiety of the BCR-ABL gene has an even chance of being the paternal or the maternal copy. We conclude therefore that there is no parental bias in the origin of the translocated ABL gene and no evidence for genomic imprinting of ABL in CML.

Splenic B cell lymphoma with "villous" lymphocytes in the peripheral blood: a disorder distinct from hairy cell leukemia.

The clinical and laboratory features of 23 patients with a characteristic form of splenic lymphoma with circulating "villous" lymphocytes (SLVL) are described and compared with those of other B cell disorders with preferential splenic involvement. SLVL affects predominantly men in their early 70's and is characterized by gross splenomegaly with little or no lymphadenopathy, presence of monoclonal gammopathy in two thirds of the cases, and infiltration of the peripheral blood by lymphocytes with a characteristic pattern of membrane irregularity. These lymphocytes are often confused with cells from hairy cell leukemia, from which they can be distinguished by a number of morphological features, by having a small cell volume, and by lack of expression of the HC2 and Tac antigens. The bone marrow is easily aspirated in the majority of cases and shows a relatively sparse infiltration. The spleen histology shows predominant white pulp involvement, although infiltration of the red pulp is seen in a small number of cases. The differential diagnosis of SLVL should also include hairy cell leukemia variant, prolymphocytic leukemia, and atypical forms of chronic lymphocytic leukemia.

An optimized multiplex polymerase chain reaction (PCR) for detection of BCR-ABL fusion mRNAs in haematological disorders.

A rapid and simple polymerase chain reaction (PCR) method is described that is capable of identifying any of the BCR-ABL transcripts that have yet been described in chronic myeloid or acute lymphoblastic leukaemia. Randomly primed cDNA is synthesized from leucocyte RNA and amplified in a single reaction containing four oligonucleotide primers (multiplex PCR). Different size products are generated from ela2 (p190) and b3a2 or b2a2 (p210) BCR-ABL transcripts which are readily and unambiguously distinguishable after agarose gel electrophoresis without the need for either nested PCR or hybridization. Chronic myeloid leukaemia cells are readily detectable even when diluted 1 in 1000 with normal blood. Samples which do not have BCR-ABL rearrangements produce a single band derived from the normal BCR gene, and the presence of this band controls for adequate RNA and cDNA preparation. Using this assay we have detected BCR-ABL transcripts in a variety of haematological disorders.

P190BCR-ABL chronic myeloid leukaemia: the missing link with chronic myelomonocytic leukaemia?

Two-thirds of patients with Philadelphia (Ph) chromosome-positive acute lymphoblastic leukaemia (ALL) have a breakpoint in the minor breakpoint cluster region (m-bcr) of the BCR gene, which results in an e1a2 transcript and a P190BCR-ABL fusion protein. This type of genomic rearrangement occurs very rarely in chronic myeloid leukaemia (CML); it has been reported in only four cases. We describe here a fifth case of P190 CML in which the cytomorphological characteristics were intermediate between CML and chronic myelomonocytic leukaemia (CMML). This case, and the four reported previously, had a consistent and significant monocytosis with a low neutrophil/monocyte ratio in the peripheral blood, resembling CMML. On the other hand, they also had a high percentage of circulating immature granulocytes, basophilia and low neutrophil alkaline phosphatase (NAP) score, which are more commonly found in classical CML. Thus, P190 CML may be a specific form of CML, in which the myeloproliferative process includes the monocytic, as well as the granulocytic lineage. Since the molecular defect in CML is thought to involve a pluripotent stem cell, the different effects of P210BCR-ABL and P190BCR-ABL in CML must reflect the somewhat wider spectrum of activity of the P190BCR-ABL. Other patients with atypical CML or CMML who lack a Ph chromosome may also have an m-bcr breakpoint which would not be detected on standard Southern blots, but which would be detectable by polymerase chain reaction amplification of reverse transcribed RNA.

Expression of the LH2 gene in chronic myeloid leukaemia cells.

The LH2 gene encodes a putative transcription factor containing two N-terminal LIM and one C-terminal HOX domains. The LH2 locus was mapped to 9q33-34.1, centromeric to the ABL gene. In a recent report, it was suggested that high levels of LH2 expression are consistently observed in chronic myeloid leukemia (CML) patients, whereas no transcription is detected in normal individuals. This led to the hypothesis that aberrant expression of LH2 may represent an additional mechanism for malignant cell proliferation in CML. We have studied the expression of LH2 in leucocytes from patients with CML or with other chronic myeloproliferative disorders (CMD), and from normal individuals, using an optimised reverse-transcription and polymerase chain reaction (PCR) technique. Twenty-seven out of 29 cDNA samples from normal individuals (93%), 49 out of 51 samples from CML patients (96%) and 20 out of 20 from Philadelphia chromosome-negative CMD showed evidence of LH2 expression. Similarly, LH2 transcription was also detected in leucocytes from CML patients in complete cytogenetic remission after treatment with interferon-alpha. Furthermore, all 36 EBV-induced lymphoblastoid cell lines established from six chronic phase CML patients showed unequivocal LH2 expression, regardless of the BCR-ABL status of the line (9 BCR-ABL positive, 27 BCR-ABL negative). We conclude that LH2 expression is not confined to CML cells, and that the t(9;22)(q34;qll) does not promote 'de novo' transcriptional activation of this gene.

Reverse transcription/polymerase chain reaction (RT/PCR) amplification of very small numbers of transcripts: the risk in misinterpreting negative results.

Technical modifications of the reverse-transcription/polymerase chain reaction (RT/PCR) amplification method now permit its use to detect amplified products from as few as one abnormal cell, either isolated or mixed with a larger number of normal cells. We studied the reproducibility of such results using as targets low numbers of cells from chronic myeloid leukaemia (CML) patients and CML cell lines in quintuplicate two-step RT/PCR designed to amplify BCR-ABL sequences. When one K562 or KYO1 cell was diluted in 10(3) non-CML HL60 cells, an amplification product was obtained in each test; at greater dilutions BCR-ABL transcripts were detected erratically. Titration of cDNA synthesised from 5 x 10(7) cells from four CML patients showed that whereas positive BCR-ABL sequences could be amplified in some tests starting with as little as a 1 in 10(7) dilution of cDNA template (corresponding to 5-10 cells), the dilution threshold for reproducible amplification was around 1 to 5 in 10(5) (100-500 cells). Quantitative PCR analysis revealed that reactions from 1 in 10(7) diluted cDNA contained less than 10 BCR-ABL transcripts as the starting template. The stochastic nature of the amplification from such small numbers of transcripts was illustrated by results of 10 replicate PCR tests on cDNA from a patient expressing both b3a2 and b2a2 transcripts: dilutions of cDNA up to 1 in 10(5) yielded dual transcript amplification in all 10 tests, but the 1 in 10(7) cDNA dilution resulted in b3a2 and b2a2 products in three tests, b3a2 only in three, b2a2 only in one and no amplification in three tests. We conclude that this 'sampling effect' may yield false-negative results and thus misinterpretation of data regarding assessment of gene expression when the quantity of target material available for study is very small.

Imatinib inhibits the activation and proliferation of normal T lymphocytes in vitro.

The ABL tyrosine kinase inhibitor imatinib mesylate is highly effective in the treatment of CML and is increasingly used in the stem cell transplantation (SCT) setting. Since ABL-dependent intracellular signaling molecules are involved in T-cell activation, imatinib may affect T-cell responses in vivo, thus affecting T-cell function in CML patients, disrupting immune reconstitution after allogeneic SCT and/or impeding the graft-versus-leukemia effect. Here we demonstrate that imatinib inhibits PHA-induced proliferation of normal peripheral blood mononuclear cells at in vitro concentrations (1-5 micromol/l) representative of the pharmacological doses used therapeutically in vivo. The effect is not dependent on antigen-presenting cells because CD3/CD28-induced T-cell stimulation was similarly inhibited by imatinib. Dose-dependent inhibition of the proliferative response of purified CD8+ and CD4+ T lymphocytes to anti-CD3/CD28 was similarly observed and associated with reduction in IFN-gamma production. The inhibitory effect could not be ascribed to an increased rate of apoptosis but the expression of activation markers on CD3+ T cells was significantly reduced in the presence of imatinib (1-5 micromol/L). Inhibition of T-cell proliferation was reversible after removal of the drug from the cultures. Thus, imatinib inhibits T-cell proliferation in vitro, an effect that is APC-independent, reversible, and does not involve apoptosis induction.

Clonal instability preceding lymphoid blastic transformation of chronic myeloid leukemia.

We have sought the presence of rearrangements of the immunoglobulin heavy chain gene locus in 13 patients with chronic myeloid leukemia (CML) in lymphoid blastic transformation (L-BT) using the polymerase chain reaction (PCR). The lymphoid nature of the transformation was confirmed by immunophenotyping and/or Southern blot hybridization with a J(H) probe. Clonal rearrangements were detected in 85% of cases and two or more rearrangements were visible in 64% of informative cases. The pattern of V(H) gene family utilization revealed an apparent reduction in V(H)4 family gene usage but otherwise reflected the known proportion of each gene family in the germline repertoire. In six cases the third complementary determining regions (CDR3) of the predominant blast crisis clone/s were sequenced revealing minimal evidence of somatic mutation. No clonal changes were detected in the chronic phase leukemia cells collected more than 6 months before the onset of L-BT in three of these patients. Of the other three patients studied in chronic phase from 1 to 6 months before L-BT, two showed clonal rearrangements which differed in size from those present at L-BT. In one patient a V(H)3 to V(H)5-D(H)-J(H) substitution had occurred at least 3 months prior to L-BT. In the other patient, however, the sequence of the rearrangement present 5 months prior to L-BT was unrelated to the rearrangements at the time of L-BT indicating a pattern of clonal succession. We conclude that: (1) IgH gene rearrangements are detectable in the majority of patients with L-BT using PCR and the lymphoid lineage of blastic CML is most readily confirmed using consensus primers to the framework 3 region; (2) somatic mutation is uncommon; and (3) B lymphoid clones distinct from those identified later may be detected before overt lymphoid BT. The identification of such 'abortive' clones is evidence for clonal instability before the onset of transformation and might have prognostic value.