Lineage involvement in chronic myeloid leukaemia: comparison between MBCR/ABL+ and mBCR/ABL+ cases

Authors


Prof. Alberto Orfao, Servicio General de Citometría, Laboratorio de Hematología, Hospital Universitario Salamanca, Paseo de San Vicente, 58–182, 37007 Salamanca, Spain. E-mail: orfao@usal.es

Summary

The relationship between different Abelson/breakpoint cluster region (BCR/ABL+) gene rearrangements and the involvement of different haematopoietic cell lineages were investigated in 15 chronic myeloid leukaemia patients. Analysis of purified cell populations confirmed the involvement of the neutrophil (89%), monocytic (89%), eosinophil (88%), erythroid (100%), and CD34+ cells (100%) in virtually all patients, without differences between minor BCR/ABL+ and major BCR/ABL+ cases; BCR/ABL+ B- and natural killer (NK)-cells were detected in 43% and 31% of cases, respectively, whereas BCR/ABL+ T-cells were rare (7%). All three minor BCR/ABL+ patients showed involvement of both B- and NK-cells, which was infrequent (27%, P = 0·06 and 10%, P = 0·01) among major BCR/ABL+ cases.

Chronic myeloid leukaemia (CML) is a clonal haematopoietic disorder associated with a characteristic chromosomal translocation – the Philadelphia (Ph)-chromosome (Daley et al, 1990), which reflects the reciprocal translocation between the long arms of chromosomes 9 and 22 [t(9;22) (q34; q11.2)] and involves the Abelson (ABL) and breakpoint cluster region (BCR) genes. This translocation typically involves an early haematopoietic progenitor capable of multilineage differentiation leading to a preferential expansion of maturing neutrophil lineage cells. Despite this, the involvement of mature cells from some haematopoietic lymphoid lineages, particularly T- and natural killer (NK)-cells, remains controversial. Accordingly, while the presence of the Ph-chromosome has been reported in a small fraction of mature T- and B-lymphocytes as well as NK-cells in some CML cases (MacKinney et al, 1993; al Amin et al, 1998; Takahashi et al, 1998; Cho et al, 2000; Nakajima et al, 2002), a great variability exists regarding the frequency of involvement of these cell populations.

Despite sharing the Ph-chromosome, CML is a cytogenetically heterogeneous disease. Most CML patients have BCR rearrangements involving the major BCR breakpoint region (MBCR) encoding for a 210 KD protein; occasionally (<1% of all CML cases) a breakpoint occurs in the minor BCR (mBCR) region (Primo et al, 2003) leading to the expression of a 190 KD BCR/ABL chimeric protein. Interestingly, despite its rarity, mBCR CML has been associated with unique clinical features frequently mimicking chronic myelomonocytic leukaemia and a higher frequency of B-lymphoid blast crisis, than MBCR cases (Hur et al, 2002).

In this study, we investigated the pattern of involvement of the different myeloid and lymphoid haematopoietic cell populations in MBCR+ and mBCR+ CML patients.

Material and methods

Patients and samples

Bone marrow (BM; n = 12) and peripheral blood (PB; n = 3) samples from 15 patients (eight males and seven females; mean age 59 ± 15 years; range 25–75 years) diagnosed with CML who were selected from a series of 230 consecutive CML cases, were studied. In nine cases, diagnostic samples were analysed while follow-up samples, obtained after therapy, were studied in the other six patients. From these latter patients, five corresponded to imatinib-resistant CML cases and one to a CML patient treated with hydroxyurea who showed no additionally acquired cytogenetic abnormalities.

Fluorescence activated cell sorting of PB and BM cell populations

Fluorescence activated cell sorting (FACS)-sorting of specific BM and PB-cell populations was performed using a FACSAria flow cytometer and the facsdiva software (Becton Dickinson Biosciences, (BDB), San Jose, CA, USA), after staining with previously described methods (Ciudad et al, 1998) for the following 4-colour combinations of fluorochrome-conjugated monoclonal antibodies (MAb)- fluorescein isothiocyanate (FITC)/phycoerythrin (PE)/peridinin chlorophyll protein (PerCP)/allophycocyanin (APC) purchased from BDB: CD7/CD56/CD19/CD3 and CD34/CD33/CD45/CD14. In two cases the CD7/CD33/CD19/CD34 staining was also used. The following cell populations were sorted: T cells (SSClow /CD3+ /CD7+), NK-cells (SSClow/CD56+/CD3/CD7+), B-cells (SSClow/CD19+), total haematopoietic stem cell and progenitor cells (HPC) (SSClow/CD45+/CD34+), neutrophil-lineage cells (SSChigh/CD45+/CD33+), monocytic cells (SSCint/CD45+/CD14+/CD33high), eosinophils (SSChigh/autofluorescent cells) and erythroid cells (SSClow/CD45). In addition, CD34+ /CD19+ B-cell precursors and CD34+/CD7+/CD33+/CD19 aberrant HPC were sorted when present at detectable levels (n = 3 and n = 2 respectively). The purity of the sorted cell populations was of 97% ± 3% (range 90–100%).

Fluorescence in situ hybridisation studies

Fluorescence in situ hybridisation (FISH) analyses for the t(9;22) translocation were performed on interphase nuclei from both erythrocyte-lysed whole BM and PB samples as well as from FACS-purified cell populations fixed in 3/1 methanol/acetic (v/v), using the LSI-it bcr/abl ES dual colour probe (Vysis Inc., Downers Grove, IL, USA). Interpretation of interphase FISH patterns was preformed according to previously described criteria (Primo et al, 2003).

Molecular analysis of BCR/ABL transcripts

RNA was extracted from washed mononuclear cells by the guanidium thiocyanate method. Reverse transcription-polymerase chain reaction analysis of the fusion gene transcripts was carried out as described elsewhere (van Dongen et al, 1999).

Results

All CML cases studied showed BCR/ABL gene rearrangements with the proportion of BCR/ABL+ cells ranging from 93% to 96% [mean ± standard deviation (SD): 95% ± 1%] in diagnostic samples and from 22% to 85% (mean ± 1 SD: 60% ± 21%) in treated cases. Overall, 12 patients had an MBCR/ABL translocation with p210 transcripts and three had p190 transcripts with an mBCR breakpoint. Within the former group, two cases displayed deletion of der(9q) and two showed a supernumerary Ph-chromosome (Table I).

Table I.  Chronic myeloid leukaemia patients (n = 15): clinical and laboratory data at the moment of entering this study together with the distribution of BCR/ABL+ cells within the different cell subpopulations.
Case no.Age (years)SexStage of the diseaseMonths after diagnosisTherapySampleBCR/ABL transcriptFISH pattern% of BCR/ABL+ cells*
Whole sampleCD34+ HPCNeutrophilsMonocytic cellsEosinophilsErythroblastsB-cellsT-cellsNK-cells
  1. NA, not analysed; BM, bone marrow; PB, peripheral blood; HU, hydroxyurea; HPC, haematopoietic stem and progenitor cells.

  2. *Results expressed as percentage of BCR/ABL+ nuclei by interphase FISH from the whole sample studied or from each FACS-purified cell population.

  3. †Percentage of BCR/ABL+ CD7+ /CD33+ /CD34+ cells of 98% and % of BCR/ABL+ CD19+ /CD34+ cells of 8%.

  4. ‡Percentage of BCR/ABL+ CD19+ /CD34+ cells of 96%.

  5. §Percentage of BCR/ABL+ CD7+ /CD33+ /CD34+ cells of 93% and % of BCR/ABL+ CD19+ /CD34+ cells of 0%.

  6. Involvement of the neutrophil, monocytic and eosinophilic lineages was systematically observed, except in case 15 who relapsed with a B-cell precursor acute lymphoblastic leukaemia.

  7. Note that in all CML cases with p190 transcripts both B- and NK-cells were always BCR/ABL+ positive, while MBCR/ABL+ cases only showed B- and NK-cell lineage involvement in 27% and 10% of the cases, respectively.

125FemaleDiagnosis0NoneBMp190t(9;22), mBCR94NA‡NANANANA80081
271MaleDiagnosis0NoneBMp190t(9;22), mBCR95NANANANANA45077
363FemaleChronic phase28ImatinibBMp190t(9;22),mBCR60759091928922051
445MaleDiagnosis0NoneBMp210t(9;22), MBCR95NA9995NANA00NA
570MaleDiagnosis0NoneBMp210t(9;22), MBCR939799939798000
655FemaleDiagnosis0NonePBp210t(9;22), MBCR94821009896NA000
757FemaleDiagnosis0NoneBMp210t(9;22), MBCR96NANANANANA000
845MaleDiagnosis0NoneBMp210t(9;22), MBCR, del(9q+)94NANANANANA000
963MaleDiagnosis0NoneBMp210t(9;22), MBCR, del(9q+)95NA†100959684NANANA
1070MaleDiagnosis0NonePBp210t(9;22), MBCR95NANANANANA3500
1175FemaleChronic phase42ImatinibBMp210t(9;22), MBCR22952471028000
1275FemaleChronic phaseNAHUBMp210t(9;22), MBCR8595959896974600
1374FemaleChronic phase44ImatinibBMp210t(9;22), MBCR6089§NANANANA04970
1439MaleMyeloid blast crisis30ImatinibBMp210t(9;22), +Ph, MBCR7210099989564000
1555MaleLymphoid blast crisis37ImatinibPBp210t(9;22), +Ph, MBCR6090000NA9700

As expected, analysis of the distribution of BCR/ABL+ cells in myeloid versus lymphoid cells showed a clearly predominant involvement of the former (100% of the cases) while only 7/14 cases studied showed BCR/ABL+ lymphoid cells. Among myeloid cells, involvement of the neutrophil, monocytic, eosinophilic and erythroid lineages was systematically observed in all but one case (case 15) who relapsed with a B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) in the absence of BCR/ABL+ myeloid cells, after attaining imatinib-induced molecular remission (Table I). Regarding lymphoid cells, a more detailed analysis of the different T-, B- and NK-cell populations showed a variable frequency of clonal involvement in each of these compartments: 6/14 cases (43%), 1/14 (7%) and 4/13 patients (31%) respectively. Interestingly, all three CML cases carrying mBCR gene rearrangements displayed a significant proportion of BCR/ABL+ cells within the B- and NK-cell compartments (percentage of positive nuclei of 49% ± 29%; range: 22–80% and of 70% ± 16%; range: 51–81% respectively) (Table I). In contrast, only four of 11 (36%) MBCR CML patients studied showed BCR/ABL+ lymphoid cells, three of them corresponding to cases treated with hydroxyurea (n = 1) and imatinib (n = 2) (Table I). In these four cases, a variable pattern of involvement of lymphoid cells was observed consisting of either MBCR/ABL+ B-cells alone (n = 3) or both MBCR/ABL+ T and NK-cells (n = 1).

Discussion

It has been well established that CML is a haematopoietic stem cell disorder. However, clonal involvement of specific haematopoietic cell lineages, such as lymphoid cells, appears to be highly variable, the reasons for which remain unclear. In the present study, we explored the potential impact of the type of BCR breakpoint on the pattern of clonal involvement of different cell populations from CML patients.

Overall, our results confirmed the involvement of the neutrophil, monocytic, eosinophil, erythroid, and CD34 HPC cells in virtually all BCR/ABL+ CML patients; interestingly, no significant differences were found between mBCR/ABL+ and MBCR/ABL+ cases with regard to the pattern of involvement of myeloid cells. Regarding lymphoid cells, variable patterns of BCR/ABL involvement were detected, in line with previous reports (MacKinney et al, 1993; al Amin et al, 1998; Takahashi et al, 1998; Cho et al, 2000; Nakajima et al, 2002). A more detailed analysis of the different lymphoid populations showed that B- and NK-cells were more frequently involved, while BCR/ABL+ T-cells were only seldomly detected. Interestingly, all three mBCR/ABL+ patients showed involvement of both B- and NK-cells. In contrast, this pattern was not detected among the MBCR/ABL CML cases. Among p210 CML, B-cells were the most frequently involved lymphoid cell population, while MBCR/ABL+ NK- and T-cells were rarely found. However, it should be noted that imatinib-resistant CML patients more frequently showed BCR/ABL+ lymphoid cells. To the best of our knowledge this is the first report of an association between the type of BCR/ABL gene rearrangement and involvement of lymphoid cells in CML patients. Altogether, these results would indicate that p190 and p210 transcripts could have a different impact on the maturation of BCR/ABL cells into the B- and NK-cell lineages. In murine models, as well as in BCR/ABL+ BCP-ALL patients, p190 expression has been shown to prevent differentiation of BCR/ABL rearranged CD34+ HPC into the myeloid, T- and B-lymphoid lineages; at the same time it imposes a B-cell programme with a partial developmental blockade (Castellanos et al, 1997; Castor et al, 2005). Such observations could help to explain the higher frequency of BCP-ALL blast crisis previously reported (Hur et al, 2002) among mBCR/ABL+ in comparison to MBCR/ABL+ CML patients, as well as the association of de novo ALL and mBCR/ABL gene rearrangements (Castellanos et al, 1997). Although a limited number of cases was analysed, it should be noted that among p210 CML cases, neither del(9q) nor supernumerary Ph were clearly associated with unique patterns of involvement of the different myeloid and lymphoid cell lineages.

In summary, our results confirm the systematic involvement of the different myeloid-lineages and CD34+ HPC in CML, while lymphoid cells were affected at variables frequencies. Interestingly, mBCR/ABL gene rearrangements were associated with unique patterns of involvement of lymphoid cells affecting both the B- and NK-cell compartments.

Acknowledgements

MD Tabernero is supported by a grant from the Ministerio de Ciencia y Tecnología MCYT, (programa Ramón y Cajal), Madrid, Spain. JM Sayagues and AB Espinosa are supported by grants (02/9103 and 02/0010) from the Ministerio de Sanidad y Consumo, Madrid, Spain.

Ancillary