| Japanese Journal of Clinical Oncology | Pages |
Acute Myelogenous Leukemia with Monosomy 7, inv(3) (q21q26), Involving Activated EVI 1 Gene Occurring after a Complete Remission of Lymphoblastic Lymphoma: A Case Report
Introduction
Materials And Methods
Case Report
Diagnosis of MDS and AML
Leukemic Cells and Cytogenetics
Leukemia Cell Lines
Preparation of mRNA and Reverse Transcription Polymerase Chain Reaction (RT-PCR)
PFGE (Pulsed-Field Gel Electrophoresis) Analysis and Southern Blot Analysis
Results
Characteristics of Leukemic Cells
Cytogenetic Analysis
Expression of the EVI 1 Gene and Ribophorin I Gene in AML with Inv(3)
Discussion
Acknowledgment
References
Acute Myelogenous Leukemia with Monosomy 7, inv(3) (q21q26), Involving Activated EVI 1 Gene Occurring after a Complete Remission of Lymphoblastic Lymphoma: A Case Report
A 42-year-old female with a mediastinal tumor and massive pleural effusion was admitted to our hospital in June 1993. Biopsy revealed lymphoblastic lymphoma. She had no evidence of distant metastasis except pleural effusion. Bone marrow examination revealed a normal karyotype (46, XY). The patient had been progression-free for more than 1 year after achieving complete remission by induction, consolidation and maintenance therapy according to the standard chemotherapy and involved-field radiation for lymphoblastic lymphoma. From May 1996 progressive leukopenia and thrombocytopenia developed. The diagnosis of refractory anemia with excess of blasts (RAEB) was made. Subsequently, in November 1996, she developed acute myelogenous leukemia (AML), M4 type by FAB classification. The karyotype of MDS and AML clones involved inversion (3) (q21q26) and monosomy 7. The EVI 1 gene was examined and was proved to be rearranged and activated. This may be the first case among the therapy-related cases of MDS/AML reported whose karyotypes were followed and in which the mRNA expression of EVI 1 gene involved was directly proved in the leukemogenesis process of chemotherapy-induced secondary MDS and AML.
INTRODUCTION
Therapy-related myelodysplastic syndrome or therapy-related acute myelogenous leukemia (T-MDS/T-AML) occurs rarely after successful treatment of hematological malignancies or cancers (1-11). Alkylating anticancer drugs or topoisomerase II inhibitors are reported to induce leukemia more frequently than other chemotherapeutic agents (7,9,11,12). Among these drugs, when used as dose-intensified chemotherapy, etoposide causes acute leukemia of myelomonoblastoid lineage with an evolution of typical chromosomal breakpoints (12-15). Alkylating agents also induce leukemia accompanied by abnormalities of chromosomes 3, 5 or 7 (15-17). Among the topoisomerase II inhibitors, anthracycline used in combination with alkylating agents or in combination with radiation has also been shown to play a role in the leukemogenesis of T-MDS or T-AML (15,16,18). Recent investigations suggested the involvement of typical oncogenes that were activated by chromosomal dislocation or deletion (17). 3q21q26 syndrome is a group of AMLs that has structural abnormality on the long arm of 3, such as inv(3) (q21q26) or t(3;3), with clinical features of abnormal megakaryopoiesis and a poor prognosis (19-24). The inv(3) (q21q26) associated with MDS progressing to AML has been reported with prominent dysmegakaryocytopoiesis (19,21,24,25). Among these, EVI 1 is reported to be associated with leukemogenesis of myeloid lineage (22,23,26-28). A few reports described cases having inv(3) (q21q26) or t(3;3) (q21;q26) who had a history of radiation or exposure to anticancer agents (23,25). However, there are few cases of T-AML in which the clinical course and karyotype changes were closely observed during the process of leukemogenesis. We report a case of lymphoblastic lymphoma who achieved continuous complete remission by intensive combined therapeutic modalities but later developed secondary MDS which transformed into AML. The evolution of the karyotypic changes with abnormalities of chromosomes 3 and 7 was closely followed and the mRNA expression of EVI 1 in the leukemic cells was proved.
MATERIALS AND METHODS
Case Report
On June 22, 1993, a 48-year-old female was admitted to our hospital because of a recent onset of back pain, persistent cough, dyspnea and low-grade fever. She had no history of cancer and there had been no history of malignancy in her family. On physical examination, the left supraclavicular lymph node was palpable and fingertip-sized. A large anterior mediastinal mass and a massive pleural effusion in the left pleural cavity were evident on X-ray-films and computerized tomograms. A sample of tumor tissue was obtained by needle aspiration and pathological diagnosis of `invasive thymoma' was tentatively made and the initial chemotherapy consisted of cisplatin (20 mg/m2 × 4 days) and adriamycin (40 mg/m2) was administered from July 2. The immune staining of mediastinal tumor tissue was positive for CD45 (LCA; leukocyte-common antigen) and CD45RO (Fig. 2 and Table 2). Systemic staging confirmed the diagnosis of lymphoblastic lymphoma, stage IIIE. Laboratory findings were serologically CRP negative and HTLV-I, HIV and HCV negative. Results of routine blood chemistry studies were within the normal range. A complete blood count (CBC) on June 23, 1993 revealed the following: WBC 8900/mm3 with 80% neutrophils, 14% lymphocytes, 4% monocytes and 2% eosinophils; 0.5% reticulocytes; hemoglobin, 12.1 g/dl; and platelet count, 381 × 103/mm3. Thus, hematological data showed no abnormality. Bone marrow (BM) examination on August 3, 1993 revealed normal cellularity, normal myeloid and erythroid precursors but lymphoma cell infiltration was present with less than 5% in total. Remission-induction chemotherapy was administered from July 28, 1993. Rapid regression of the large mediastinal mass and a decrease of the left pleural effusion were observed. In January 1994, uncertain complete response (CRu) was achieved. In November 1994, regrowth of anterior mediastinal mass occurred and recurrence of the lymphoblastic lymphoma in BM developed. The patient was administered reinduction combination chemotherapy, resulting in shrinkage of the mediastinal mass. From February 1995, involved-field radiation therapy for the residual mass was performed with a total of 60 Gy and CRu was again achieved in April 1995. No cytotoxic drugs were administered further. For more than 1 year, she remained as an uncertain CR, with no signs of recurrence. Chemotherapy and radiation therapy are summarized in Table 2.
Beginning on May 13, 1996, progressive leukopenia and thrombocytopenia developed. The WBC count was 2700/mm3 with 58% neutrophils, 28% lymphocytes, 13% monocytes, 1% eosinophils. Hemoglobin was 12.4 g/dl and the platelet count was 117 × 103/mm3. BM examination was performed on July 12, 1996 and revealed normocellular marrow, decreased megakaryocyte numbers and a dysmyelopoiesis with a megaloblastoid change and increased numbers of blast cells; i.e. 14% blast cells, 47% myeloid series, 2.5% eosinophils, 1.5% monocytes, 23% erythroblasts and 12% lymphoid cells. Blast cells showed a myelomonoblastoid origin. No lymphoma cells were detected in the BM. The diagnosis of refractory anemia with excess of blasts (RAEB) was made. Examination of BM on November 21, 1996 revealed an increase in myelomonoblastoid cells to 31.4% and eosinophils 2.5% and CBC revealed a WBC count of 5800/mm3 with 25% blastoid cells and 0% eosinophils. Hemoglobin was 9.7 g/dl and the platelet count was 63 × 103/mm3. The diagnosis of AML (FAB M4) was made which transformed from RAEB. Induction chemotherapy consisting of a 3-day administration of idarubicin (7 mg/m2) and 7-day administration of cytosine arabinoside (100 mg/m2) was initiated early in January. The induction treatment failed and no aggressive chemotherapy was administered further because of the manifested anthracycline-induced late cardiac complication. The patient was followed by the best supportive care and died of progressive disease in April 1997.
Diagnosis of MDS and AML
Hematological diagnosis was made by morphological and cytochemical studies of peripheral blood and BM smears. The FAB (French-American-British) Cooperative Group criteria were used (29). Immunophenotypes of leukemic cells were analyzed by flow cytometry study using routine monoclonal antibodies.
Leukemic Cells and Cytogenetics
The patient's bone marrow (BM) cells and peripheral blood leukemic cells were obtained for research analysis after informed consent. Cytogenetic analysis was by routine G-banding using a Giemsa staining technique. Metaphase cells were examined from direct preparations and short-term culture (24-72 h). The karyotypes were described according to the International System for Cytogenetic Nomenclature (30).
Leukemia Cell Lines
UCSD/AML 1 cell line (31) with t(3;3), Kasumi-4 (32) with inv(3) and the MOLT16 cell line (33) which was originally established from T-cell ALL were used in the experiments. The UT-7 cell line was kindly provided by Dr Komatsu (Jichi Medical School) and was used as a control for PFGE as described later.
Preparation of mRNA and Reverse Transcription Polymerase Chain Reaction (RT-PCR)
BM cells and peripheral blood cells were obtained in December 1996 before starting treatment of AML. Poly(A)+RNA from frozen cells was extracted with a Fast Track mRNA Isolation Kit (Invitrogen) according to the manufacturer's instructions. One microgram of each poly(A)+RNA sample was reverse transcribed to cDNA primed with oligo(dT) in a 20 µl reaction volume with 200 U of Moloney Murine Leukemia Virus reverse transcriptase (Life Technologies). PCR was performed on a portion (0.5 µl) of each cDNA mixture in a 20 µl reaction volume containing 25 pmol of each primer, 5 U of Taq DNA polymerase (Pharmacia Biotech), 0.2 mM of each deoxynucleoside triphosphate, 1.5 mM MgCl2 and 1× reaction buffer (Pharmacia Biotech). After 30 rounds of PCR (30 s at 94°C, 30 s at 57°C, 1 min at 72°C), 10 µl of PCR product were electrophoresed in a 2% agarose gel. The primers used were as follows:
EVI 1F1:5[prime]-CGAAAGCGAGAATGATCTCC-3[prime]EVI 1R2:5[prime]-GGAAGACGTAGTGCTGAACA-3[prime]RBPNF1:5[prime]-GTTTCTGCTCCTGTTGCTTG-3[prime]RBPNR1:5[prime]-AAGAGCAACTGGGAGCTTGA-3[prime]PFGE (Pulsed-Field Gel Electrophoresis) Analysis and Southern Blot Analysis
A
 B  |
Figure 1. (A) Hematoxylin-eosin (H-E) staining of lymphoid cell infiltration of biopsied mediastinal mass (June 24, 1993). Immune staining of tumor cells exhibited CD45 (+), CD45RO (+) and CD20 (-). Original size ×80. (B) H-E staining of left neck lymph node (July 2, 1993). The tumor consisted of proliferation of medium-sized lymphocytes with starry sky appearance, and among these a few elements of epithelioid cells showed microcystic space structure. Immune-staining showed CD3 (-), CD4 (-), CD8 (-), CD19 (-), CD20 (-), S-100 (-) and cytokeratin (-). Original size ×160.
Agarose plugs from patient cells or from UT-7/GM cells as a control were prepared according to the method described by Morishita et al. (27). Briefly, DNAs in the plugs were digested with the indicated enzymes overnight. After digestion, the plugs were loaded on to a 1% agarose gel. PFGE was performed on the CHEF Mapper apparatus (Bio-Rad, Richmond, CA) at 14°C with ramped pulses from 20 to 90 s for 19 h at a constant voltage of 200 V. After PFGE, the DNAs were transferred to charged nylon filters (Pall BioSupport, East Hills, NY). The filters were hybridized as described previously (34). The H1-1 probe is a 2.4 kilobase (kb) EcoRI cDNA fragment that contains the EVI 1 gene coding sequence, located on 3q26. The 27A Hd/Apa probe is a 0.8 kb HindIII-ApaI genomic DNA fragment from a cosmid DNA, located on 3q21 (23).
RESULTS
Characteristics of Leukemic Cells
The bastoid cells in the BM of the MDS and the leukemic cells of the AML transformed from MDS were positive for peroxidase, [alpha]-naphthol ASD-chloroacetate and [alpha]-naphthyl butyrate esterase (Fig. 2A and 2B). Immunophenotypes of leukemic cells (December 19, 1996) showed positive for CD11b (25.1%), CD11c (39.5%), CD13 (79%), CD14 (15.2%), CD33 (60.9%), CD34 (63.4%), CD38 (60.6%), anti-HLA-DR (79.2%) and MPO (80.1%) as shown in Table 1.
A
 B  |
Figure 2. (A) May-Giemsa staining of bone marrow smear (June 24, 1996), which contained 14% of myelomonoblastoid cells. Original size ×400. (B) May-Giemsa staining of leukemic cells in the bone marrow (December 16, 1996), which showed double positive in esterase-double staining including [alpha]-naphthol-ASD-chloroacetate esterase and [alpha]-naphthyl-butyrate esterase. Original size ×400.
Table 1.
| Cytochemical | Immunophenotypes | |
| Mediastinal tumor (June 24, 1993) | N.T. | CD3 (-), CD4 (-), CD8 (-), OKT6 (-), CD20 (-), S-100 (-), cytokeratin (-), CD45 (+), CD45RO (+) |
| Leukemic cells (December 19, 1996) | Peroxidase (+) [alpha]-Naphthol ASD chloroacetate esterase (+) [alpha]-Naphthyl butyrate esterase (+) | CD2 (-), CD10 (-), CD19 (-), CD11b (+), CD11c (+), CD13 (+), CD14 (+), CD33 (+), CD34 (+), MPO (+) |
Table 2.
| 1993 | 1994 | 1995 | Total | |
| Cisplatin (mg/m2) | 80 | 80 | ||
| Vincristine (mg/m2) | 15.4 | 7 | 22.4 | |
| Cyclophosphamide (g/m2) | 3.6 | 1.19 | 4.79 | |
| Doxorubicin (mg/m2) | 320 | 40 | 360 | |
| Daunorubicin (mg/m2) | 120 | 120 | ||
| Mitxantron (mg/m2) | 18 | 18 | 36 | |
| Etoposide (mg/m2) | 500 | 720 | 1220 | |
| l-Asparaginase (U/m2) | 60000 | 60000 | ||
| Cytosine arabinoside (mg/m2) | 480 | 480 | ||
| Vindesine (mg/m2) | 2.4 | 5 | 7.4 | |
| ACNU (mg/m2) | 120 | 120 | ||
| Methotrexate (mg/m2) | 50 | 1050 | 1100 | |
| 6-Mercaptopurine (g/m2) | 5.88 | 5.88 | ||
| Radiation (mediastinum; Gy) | 60 | 60 |
Table 3.
| Date analyzed | Karyotype* | Cells analyzed |
| October 19, 1993 | 46, XX | (20/20) |
| July 2, 1996 | 46, XX, inv(3) (q21q26) | (10/20) |
| 45, XX, inv(3) (q21q26), -7 | (7/20) | |
| 46, XX | (3/20) | |
| November 26, 1996 | 45, XX, inv(3) (q21q26), -7 | (19/20) |
| 46, XX | (1/20) | |
| January 8, 1997 | 45, XX, inv(3) (q21q26), -7 | (18/20) |
| 46, XX | (2/20) |
Cytogenetic Analysis
Chromosomal examination of BM on October 7, 1993 revealed normal karyotype of 46, XX as shown in Table 3. However, BM cells on July 6, 1996 when the MDS was diagnosed revealed two clones containing inversion 3 alone or monosomy 7 and inversion 3. After transformation to the AML, karyotype analysis showed 45, XX, inv(3) (q21q26), -7, as shown in Fig. 3.
Figure 3. Karyotype of bone marrow leukemic cells of January 8, 1997, which showed 45, XX, inv(3) and monosomy 7. Arrows indicate inv(3) (q21q26) and loss of chromosome 7. To determine the genetic alteration by which the EVI 1 gene expressed in leukemic cells in this case, we performed pulsed-field gel electrophoresis (PFGE) followed by Southern blot hybridization analyses with specific probes, H1-1 and 27A Hd/Apa, located on 3q26 and 3q21, respectively. As shown in Fig. 4, using the H1-1 probe, we detected each rearranged band in all digested DNAs in addition to a germinal band. Likewise, using the 27A Hd/Apa probe, we also detected rearranged bands. The rearranged bands recognized by independent hybridization were identical between the two probes. As SfiI and BssHII(NruI) fragments were rearranged with the H1-1 probe, it was suggested that the breakpoint on 3q26 existed within the BssHII(Nru I)-SfiI [sim]520 kb fragment containing the EVI 1 locus indicated in Fig. 6. However, we detected no rearranged fragments with EcoRI, BamHI and HindIII digestion by the usual genomic DNA Southern blot hybridization analysis with EVI 1 cDNA probe (data not shown). Therefore, the results suggested that the breakpoint was presumed to occur at 3[prime] of the EVI 1 gene, outside the coding region. These results are consistent with our previous mapping of inv(3) cases to a region at 3[prime] of the EVI 1 gene (23). On the other hand, as BssHIII and NotI fragments were rearranged with the 27A Hd/Apa probe, the breakpoint on 3q21 was predicted to be located within the [sim]50 kb NotI fragment in which breakpoints concentrated as shown in previous reports (23,34). On additional genomic DNA Southern blot hybridization analyses with the 27A Hd/Apa probe, we detected a rearranged band (data not shown), so we concluded that the breakpoint on 3q21 was located in the breakpoint cluster region involved in 3q21q26 syndrome, designated in our previous study (23).
Figure 4. Detection of the chromosomal rearrangements at 3q21 and 3q26 by PFGE followed by Southern blot hybridization analyses. High molecular weight DNAs from leukemic cells (P) from this case and control cells (C) were each digested with SfiI, BssHII, NruI and NotI. The UT-7/GM cell line was used as control cells. The 27A Hd/Apa and H1-1 probes were used to detect the chromosomal rearrangement at 3q21 (A) and 3q26 (B), respectively. Molecular sizes of DNAs are shown on the left side of the figures. Asterisks indicate identical rearranged bands between 3q21 and 3q26. To determine the expression of the EVI 1 gene and Ribophorin I genes, we performed RT-PCR analysis. The UCSD/AML1 cell line and Kasumi-4 cell line were used as positive controls of the EVI 1 gene and Ribophorin I gene expression, respectively, according to a previous report (23). The MOLT-16 cell line was used as a negative control of the EVI 1 gene expression. As shown in Fig. 5, the leukemic cells of this AML case with genetic alteration, inv(3), was detected on expression of the EVI 1 gene. On the other hand, the Ribophorin I gene was expressed in this patient's leukemic cells as well as other cell lines. Figure 5. Identification of the EVI 1 and Ribophorin I gene transcripts by RT-PCR. Poly(A)+RNA from frozen cells of UCSD/AML (lane 1), Kasumi-4 (lane 2), the patient's peripheral blood cells (lane 3) and MOLT-16 (lane 4) were used as each cDNA template. Lane 5 was loaded with the PCR product which contained no cDNA template. GAPDH gene transcriptions show complete reverse transcription from each poly(A)+RNA. Figure 6. Physical map of 3q21 and 3q26 implicated inv(3) and t(3;3). White boxes indicate location of the genes on chromosome and the positions of these genes were established in previous studies (23,34). The chromosomal breakpoints in other cases of AML or CML-BC with inv(3) (q21q26) (t) or with t(3;3) (q21q26) ([darr]) are indicated. The predicted chromosomal breakpoint in this case is indicated by an arrowhead. Table 4. 
Expression of the EVI 1 Gene and Ribophorin I Gene in AML with Inv(3)
A
B
Ref.
Chromosome abnormalities
AML diagnosis and clinical features
22
47, XX, inv(3) (q21q26), del(7) (q22)
de novo M1 with thrombocytosis
23
46, XY, inv(3) (q21q26),
MDS converted to M0
46, XY, inv(3) (q21q26),
de novo M2 with thrombocytosis
46, XX, inv(3) (q21q26), -7
de novo M7 with thrombocytosis
46, XY, inv(3) (q21q26), -7
T-MDS/T-AML secondary to neuroblastoma
26
46, XY, inv(3) (q21q26)
MDS converted to M6
46, XY, inv(3) (q21q26), iso7(q)
MDS converted to M7
46, XY, inv(3) (q21q26),
de novo M4
46, XY, inv(3) (q21q26) × 2, t(9;17;22)
CML blastic crisis (M4)
46, XY, inv(3) (q21q26),
de novo M5
This case
45, XY, inv(3) (q21q26), -7
T-MDS/T-AML secondary to lymphoblastic lymphoma
DISCUSSION
There have been many reports of the development of AML after an initial diagnosis of T-lymphoblastic lymphoma (35,36). It is well known that neoplastic cells in some patients with T-lymphoblastic lymphoma express myeloid antigens in addition to T-lymphoid antigens. Since both T-lymphoblastic lymphoma and acute lymphoblastic leukemia frequently express myeloid antigens, secondary leukemia must be differentiated from a biphenotypic hematologic malignancy or an acute mixed-lineage leukemia and also the possibility of malignant transformation of pluripotent progenitor cells capable of differentiation in either the myeloid and lymphoid neoplasm must be ruled out (37). Morphological studies of the initial lymph node in this case did not show immature granulated cells suggestive of myeloblasts and the initial bone marrow examination did not show myeloid hyperplasia or dysplasia. In addition, in this case, a CR of lymphoblastic lymphoma was achieved for more than 18 months and the AML cells that later developed did not express lymphoid-associated antigens (Table 1). Regarding the possibility of a simple phenotypic change from T-lymphoblastic lymphoma to AML in an identical neoplastic clone, recently a syndrome of lymphoblastic lymphoma, eosinophilia and myeloid hyperplasia/malignancy associated with t(8;13) (p11;q11) was described by Abruzzo et al. (36). However, this case differs, in that no eosinophilia in the initial examination of lymphoma tissue or in the initial bone marrow smear was detected and the cytogenetic analysis of the leukemic cells did not show t(8;13) (p11;q11). In addition, the development of AML occurred 3 years after the initial diagnosis of lymphoma, whereas most reported cases of this biphenotypic neoplastic syndrome arose within 1 year of diagnosis and did not show any preceding myelodysplastic syndrome. The expression of CD11b, CD11c, CD13, CD14, CD33 and MPO of AML cells strongly indicates that the cells were committed to myelomonocytic lineage and not to the lymphoid lineage and were not of a biphenotypic type. In contrast, AML cells lacked the expression of CD2(pan-T) or CD19(pan-B). Therefore, this case is strongly suggestive of therapy-related AML rather than a biphenotypic hematologic malignancy.
The incidence of T-MDS/AML was reported to increase particularly when combined modalities including radiation were used for treatment (4,6-9,13,16,38,39). Standard therapy for lymphoblastic lymphoma or acute lymphoblastic leukemia has been established based on the results of L10 and L17 protocols or those of the German ALL Study Group (40,41). In our case, the induction regimen consisted of doxorubicin, vincristine, cyclophosphamide, prednisolone and l-asparaginase and the consolidation chemotherapy and the intensive chemotherapy consisted of non-cross-resistant anticancer drugs, such as etoposide, cytosine arabinoside, methotrexate and ACNU. The delivered radiation dose was 60 Gy in terms of involved field radiation for the residual tumor which mimicked `thymic cancer' after salvage chemotherapy. The therapeutic drugs and dosages are shown in Table 2. The total doses of each anticancer drug except doxorubicin were almost within the upper limits planned. The chemotherapy regimens and the radiation dose were also similar to the standard protocol of L10 or L17 (41). The patient achieved CR for more than 1 year after successful salvage therapy using combined modality, but developed T-MDS/T-AML.
Risk factors for the development of T-MDS/T-AML are reported to be as follows: (i) prolonged exposure to alkylating agents, (ii) intensive treatment by epipodophyllotoxins, (iii) combined use of alkylating agents and radiation, (iv) radiation to the pelvic bone (BM-rich area), (v) asplenism and (vi) being over 40 years old (11,12,14,16). BM stem cells that are damaged intrinsically by alkylating agents are likely to contribute to the development of T-MDS/T-AML, especially after autologous stem cell transplantation (14). Bitter et al. (19) noted that nearly half of patients with abnormalities of both chromosomes 5 and 7 had been treated with combined modality therapy. In our case, the time from the diagnosis and therapy of cancer to the onset of myelodysplastic syndrome was 34 months. The evolution of the clonal abnormality was expressed as a chromosomal abnormality accompanied by the development of MDS. The region of chromosomes 5, 7 or 11 observed in T-MDS/T-AML represents a target for mutagenic agents that might cause loss of genes on these chromosomes and/or intrachromosome relocation. This provides support for the idea that following the loss or deletion of genes, other genes could be inactivated or activated, resulting in abnormal cell growth. Thus, the common chromosomal abnormalities observed in the T-MDS/T-AML associated with therapeutic exposure to alkylating agents are 5q- and monosomy 7, whereas those observed in T-AML associated with topoisomerase II-reactive drugs, i.e. epipodophyllotoxins or anthracyclines, are a genetic abnormality at 11q23 or translocations involving 21q22 (15,18). On the other hand, the specific causative agents inducing the abnormality at 3q26 and its association with leukemogenesis remain to be determined (19,21). The proposed mechanisms of iatrogenic leukemogenesis caused by topoisomerase II inhibitors usually occur whereby low (under cytocidal) concentrations of the cytotoxic agent triggers the genesis of abnormal hematopoietic clones (42). These drugs at cytocidal concentrations usually induce chromosomal fragmentation, sister chromatid exchange, DNA rearrangements and apoptosis at the cellular level. Such drugs produce a covalent topoisomerase-DNA cleavable intermediate complex and increase the risk of T-MDS/T-AML (16,42).
3q21q26 syndrome is among a group of MDS and AML which have structural abnormality on the long arm of 3, such as inv(3) (q21q26) or t(3;3), with clinical features of abnormal megakaryopoiesis and a poor prognosis (19,20,24,43) (Table 1). However, we could not detect evidence of dysmegakaryopoiesis, such as an elevated platelet count. Fonatsch et al. (25) reported 18 cases of inv(3) (q21q26) or t(3;3) (q21q26), of which three had a history of exposure to radiation or to chemotherapeutic agents. They postulated that 3q21q26 syndrome was characterized by a mutagen exposure and a symptomatic myelodysplasia before transformation to AML, which is consistent with our case. In terms of the abnormal clone, eight out of 18 cases in their report had co-existence of inv(3) (q21q26) and monosomy 7 in the same clone. In our case, initially two separate clones of inv(3) (q21q26) alone or inv(3) (q21q26) and monosomy 7 were detected. A relatively high frequency of T-MDS/T-AML has been reported by Bitter et al. (19) and Jenkins et al. (20). Therefore, it may be concluded that 3q21q26 syndrome with or without the chromosome 5 or 7 abnormality occurs relatively frequently and represents a secondary leukemia. The dual breakpoints at 3q leads to the juxtaposition of the DNA sequence at 3q21 and 3q26 and to the activation of zinc finger genes related to EVI 1. Some abnormality of inv(3) was also reported in the M7 type of AML(23,26). EVI 1 of 3q26 is reported to be activated in these de novo AML (26). We observed the transcriptional activation of the EVI 1 gene, which is mapped to 3q26, in the leukemic cells of our case. Although Southern blot analysis using probes inside the amplified fragments is required for presenting data obtained by the RT-PCR technique, we usually confirm the sequence of the bands detected by the primers in RT-PCR. Based on our previous observations, EVI 1 and Ribophorin I gene transcripts in the leukemic cells of this patient could be identified (23). The clusters of breakpoints of 3q26 in most cases of inv(3) occur downstream of the 3[prime] region of the EVI 1 gene, which is consistent with our case. Regarding the results obtained by PFGE/Southern blot analysis, we observed the additional bands in the genomic DNA of leukemic cells. It appeared that these additional bands (without asterisks in Fig. 4) in the control lanes were present and they disappeared in the lanes of the patient's samples. The reason why these additional bands appeared may be due to technical considerations, that is, the amount of DNA (loaded plug) on the control lanes and the patient's samples differed, causing a shift of germ line bands in the control lanes. However, these additional bands were detected in almost all data obtained by using the restriction enzymes, including SfiI, which is essentially insensitive to methylation. In addition, these additional bands were reproduced similarly and we could detect similar germ line bands with a shift when we used the same membrane and performed a hybridization experiment with the probes obtained from another chromosome such as chromosome 1. Therefore, we consider these bands to be identical rearranged bands originating from chromosomal translocations. Based on these results we consider that the breakpoint of 3q21 is present within 50 kb near NotI including 27A Hd/Apa, that is, downstream of the 3[prime] region of Ribophorin I as shown in Fig. 6. Recent studies suggested a possible role for Alu-mediated homologous recombinations as a molecular mechanism for these leukemogenesis (42). In the eipodophyllotoxin-related leukemias, the 11q23 translocation breakpoint sites have been confirmed within the Alu sequences. However, the role of DNA topoisomerase II in homologous recombination involving Alu repeats in the leukemogenesis of 3q21q26 syndrome remains a question (18). The phenotype of AML in our case had features of the FAB, M4, different from that of secondary AML with chromosome 5 or 7 showing M1 or M2 type, whereas epipodophyllotoxin-related secondary AML shows M4 type (12). The overall prognosis of T-MDS/T-AML patients is poor and an estimated 12 month survival rate of 10% has been reported despite salvage therapy (39,43). It has also been reported the median overall survival was actually longer in patients who had karyotypes associated with a poor prognosis and received supportive care only and who did not receive cytotoxic treatment for T-MDS/T-AML. Other treatment approaches must be established for those patients with T-MDS/T-AML who respond poorly.
In conclusion, we have reported a case of myelodysplastic syndrome secondary to chemotherapy and of lymphoblastic lymphoma which converted to AML (M4). The karyotype of MDS and AML clones involved inv(3) (q21q26) and monosomy 7. The EVI 1 gene was rearranged and transcriptionally activated. The chromosomal breakpoints at 3q21 and 3q26 were identified in the translocations. To our knowledge, this case was the first among T-MDS/T-AML whose karyotypes were followed and the activation of the EVI 1 gene involved was directly proved in the process of leukemogenesis of chemotherapy-induced secondary MDS and AML. These results support the hypothesis that activation of the expression of the EVI 1 gene may be involved in T-MDS and leukemogenesis.
Acknowledgment
This study was supported by a Grant-in-Aid for Cancer Research from the Ministry of Health and Welfare, Japan
References
For reprints and all correspondence: Tadahiko Igarashi, Department of Medicine, National Cancer Center Hospital East, 5-1, Kashiwanoha 6-chome, Kashiwa, Chiba, 277, Japan. E-mail: tigarash@east.ncc.go.jp
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 24 Nov 1998
Copyright©Japanese Journal of Clinical Oncology, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This Article 
Abstract
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (2)
Request Permissions
Google Scholar
Articles by Igarashi, T
Articles by Mukai, K
Search for Related Content
PubMed
PubMed Citation
Articles by Igarashi, T
Articles by Mukai, K
Social Bookmarking
What's this?