Reverse Transcriptase-Polymerase Chain Reaction for bcr/abl Fusion in Chronic Myelogenous Leukemia

Transcription

1 HEMATOPATHOLOGY Reverse Transcriptase-Polymerase Chain Reaction for bcr/abl Fusion in Chronic Myelogenous Leukemia STEPHEN J. WELLS, MD, CAROL N. PHILLIPS, MS, ELLIOT F. WINTON, MD, AND DIANE C. FARHI, MD Reverse transcriptase-polymerase chain reaction (RT-PCR) has shown promise as a means of detecting low levels of cells bearing the Philadelphia chromosome (Phi) and for detecting cytogenetically inapparent ("masked") Phi in patients with chronic myelogenous leukemia (CML). For detection by karyotyping, dividing cells must be used, precluding use of peripheral blood samples in cases with low peripheral blood blast counts. Reverse transcriptase-polymerase chain reaction was performed in 83 bone marrow and 30 peripheral blood samples from patients with CML to compare results with karyotyping and to evaluate utility of this test on peripheral blood samples. Using isolated total cellular RNA and a single primer pair, cdna was transcribed, amplified, electrophoresed, and probed for bcr/abl fusion involving M- bcr exons 2 and 3 of the bcr gene. Fifty-three samples were from untreated or conventionally treated patients (pre-bmt), and 60 were from patients who had undergone bone marrow transplantation (post-bmt). Fifty of 53 pre-bmt samples were positive by RT-PCR. Two samples, negative by RT-PCR, had complex translocations, 1(9; 16; 22) and t(4;l4;22). One case was indeterminate by RT-PCR, but positive on retesting. Forty-five of 53 had Phi by karyotyping; 8 were negative, The cytogenetic hallmark of chronic myelogenous leukemia (CML) is the Philadelphia chromosome (Phi), t(9;22)(q34;ql 1).' This reciprocal translocation is present in 90% to 95% of cases of CML by karyotypic analysis. At the molecular level the Ph 1 results from breaks in the c-abl proto-oncogene (abl) located on chromosome 9q34, and in a 5.8 kilobase area of the breakpoint cluster region (bcr) gene on chromosome 22q ll. 2,3 The bcr locus involved in the Phi in CML is often referred to as the major region of the bcr gene or M-bcr. The region of chromosomes 9 and 22 telomeric to the breaks are switched, forming a hybrid gene on chromosome 22, which combines the centromeric portion of the bcr gene From Emory University School of Medicine. Atlanta, Georgia. Manuscript received July 5, 1995; revision accepted January 4, Address reprint requests to Dr. Farhi: Department of Pathology, F- 143 Emory University Hospital, 1364 Clifton Road, NE, Atlanta, GA including 5 peripheral blood samples, 2 bone marrow samples with "masked" Phi, and 1 bone marrow sample with poor growth. Thirtyfive of 60 post-bmt samples were positive by RT-PCR. Fourteen of 60 post-bmt samples had Phi by karyotyping. Of the RT-PCR+/Phlcases, most showed a weak but definite band by RT-PCR, suggesting a low level of the bcr/abl fusion gene. Nineteen patients had concurrent peripheral blood and bone marrow samples analyzed by RT-PCR and karyotyping. Of 16 patients with satisfactory RNA extraction, 15 had concordant results by RT-PCR. Five patients had adequate metaphase cells for karyotypic analysis. All had Phi in bone marrow, but were negative in peripheral blood. Our results indicate that RT-PCR for detection of bcr/abl fusion is more sensitive than karyotyping in pre- and post-bmt samples. Furthermore, RT-PCR can be successfully performed on peripheral blood, yielding excellent correlation with bone marrow samples. (Key words: bcr/abl; Chronic myelogenous leukemia; Philadelphia chromosome; Bone marrow transplantation; Polymerase chain reaction; Reverse transcriptase) Am J Clin Pathol 1996; 105; and the telomeric portion of the abl gene. This bcr/abl fusion gene is transcribed to a chimeric bcr/abl messenger RNA, which is in turn translated to a bcr/abl fusion protein of 210 kd, which has enhanced transforming capacity in vitro. In CML the break in the abl gene usually occurs in intron 1. The break in the for gene is most often between exons 2 and 3 or between exons 3 and 4 of the M-bcr. 4 Traditional karyotyping has a long and proven record of success in detecting the Ph 1. Karyotyping may also demonstrate additional cytogenetic abnormalities, which typically portend a more aggressive clinical phase of CML. 5 However, alternate methods of detecting bcr/ abl fusion are useful in cases of suspected CML when the Phi is karyotypically inapparent or "masked." In addition, routine karyotyping may not be sufficiently sensitive to detect minimal residual disease after therapy. Furthermore, peripheral blood (PB) samples with low blast counts may not be acceptable for karyotyping, as dividing cells are necessary for analysis. 756

2 WELLS ET AL. 757 RT-PCR for bcr/abl in CML Recently, an RNA-based RT-PCR has been developed to detect bcr/abl fusion in patients with CML and acute leukemia. 6 Generally this has been done with a "nested" PCR technique, using a second set of internal primers and reamplification after one round of PCR amplification. Although nested PCR may enhance assay sensitivity, it provides additional opportunity for specimen contamination and false positive results. In this study, we performed a nonnested RT-PCR for bcr/abl fusion in bone marrow (BM) and PB samples from patients with CML. These results were compared with karyotypicfindings to evaluate the relative sensitivities of these methods. Additionally, we studied concurrent BM and PB samples by both techniques to evaluate the utility of RT-PCR and karyotyping with PB samples. MATERIALS AND METHODS One hundred and thirteen samples from patients with CML were evaluated for bcr/abl fusion by RT-PCR. These included 83 BM and 30 PB samples. Conventional karyotypic analysis was also performed on each sample studied by RT-PCR. Fifty-three samples were from untreated patients or patients treated with hydroxyurea or alpha-interferon. Cumulatively, these patients are referred to as the pre-bone marrow transplant (pre-bmt) group. The remaining 60 patients had undergone allogeneic BMT 1 to 76 months before RT-PCR evaluation (post-bmt). Nineteen patients had concurrent BM and PB samples for analysis for RT-PCR and karyotyping. Seventeen additional BM samples from patients with chronic myelomonocytic leukemia (2 patients), polycythemia vera (1 patient), essential thrombocythemia (1 patient), myelofibrosis with myeloid metaplasia (3 patients), and acute myelogenous leukemia (10 patients) were evaluated by both RT-PCR and karyotyping. RT-PCR A non-nested RT-PCR was used for the detection of bcr/abl fusion. The method is a modification of that reported by Kawasaki and colleagues. 6 Briefly, total cellular RNA was isolated and extracted using an RNAzol B (Biotecx, Houston, TX)/chloroform/alcohol method. Gel electrophoresis of a 2 ^L aliquot of patient RNA was performed and an approximate RNA concentration estimated by comparison of band intensity with a placental RNA control. A 10 nl RT cocktail was prepared consisting of 1 Mg total RNA, 3 ram MgC12, 1 X PCR buffer, 1.0 mm each of datp, dctp, dgtp and dttp, 1 U/^L RNase inhibitor (Perkin Elmer, Foster City, CA), 2.5 U/ nh MuLV reverse transcriptase (Perkin Elmer), and 0.5 um downstream primer complementary to abl exon, 2, 5TCAGACCGTGAGGCTAAAG3'. The RT mixture was incubated at 42 C for 1 hour, heat inactivated at 99 C and then cooled. The resultant cdna was combined with 40 jul bcr/abl PCR mix containing 1.8 mm MgC12, 1 X PCR buffer, 0.2 mm each of datp, dctp, dgtp and dttp, 1.25 U Amplitaq (Perkin Elmer), 0.8 mm spermidine, and 0.5 um upstream primer complementary to M-bcr exon 2, 5TGAAACTCCAGACTGTCCAC3'. The reaction mix was brought to 94 C in a thermal cycler, then cycled 40 times as follows: 40 C for 30 seconds, 55 C for 1 minute, 72 C for 1 minute, followed by a 7 minute extension at 72 C and a 4 C soak. Gel electrophoresis of 25 ;ul of each reaction product was performed, followed by Southern transfer 7 to a nylon membrane, and hybridization for 1 hour with 30 /*L of biotin end-labeled probe complementary to M-bcr exon 2, 5'ATTGATGGTCAGCGGAATGCT3'. A chemiluminescent detection system (Southern-Light Chemiluminescent System, Tropix, Bedford, MA) was employed with overnight exposure to radiographic film. With each specimen run, patient and K.562 cell line controls were used to assess breakpoints within M-bcr exons 2 and 3, respectively. A negative control sample (placental RNA) and a specimen blank were also evaluated with each specimen run. In addition, RT-PCR was performed on adilution series of 10-1 to 10-5 MgofK.562 cell line RNA diluted in placental RNA to ensure assay sensitivity. Patient specimens were run in duplicate in each case. Physically separated, dedicated working areas for specimen preparation, pre-pcr, PCR, and post-pcr analysis were used. The distinction between M-bcr breakpoints between exons 2 and 3 or between exons 3 and 4 was based on slight differences in size and electrophoretic migration of the PCR products. Comparison with the location of positive control bands was performed to determine the breakpoint site. A distinct band of at least equal intensity to the band formed by 10' 5 ng of K562 cell line RNA in an appropriate location (as compared with positive controls) was interpreted as a positive result. A very faint band of lesser intensity than that formed by 10" 5 ng of K562 cell line RNA or failure of a sample to yield duplicate identical results was considered an indeterminate test. RT-PCR was repeated in this case. Repeated very faint bands were not encountered in this study. Karyotypic analysis was performed on unstimulated BM and PB cells cultured for 24 and 48 hours in AmnioGrow HI media (Inovar, Gaithersburg, MD). Metaphases were collected by centrifugation andfixedin Vol. 105-No. 6

3 758 HEMATOPATHOLOGY three parts methanol: one part acetic acid. In general, at least 15 metaphase spreads were analyzed following trypsin-giemsa banding. In cases of poor cell growth, all available metaphases were studied. Chromosomes were classified according to the International System for Human Cytogenetic Nomenclature (ISCN). 8 RESULTS Fifty of 53 samples from pre-bmt patients were positive by RT-PCR (94%). Negative cases included two samples with complex translocations, t(9;16;22) and t(4;14;22), and one sample, which was indeterminate with a questionable very weak band, but positive on retesting. Forty-five of these 53 pre-bmt samples (85%) demonstrated t(9;22) or a variant Phi by karyotyping. The eight cases lacking Ph 1 by karyotyping included two with "masked" Phi, one sample with poor growth, and five PB samples. Each of the PB samples hadfiveor fewer metaphase cells for analysis. Thirty-five of 60 samples (58%) from post-bmt patients were positive by RT-PCR. Five samples had degraded RNA, unsuitable for RT-PCR. The remaining 20 negative samples had intact RNA, but no detectable bands. Fourteen of the 60 post-bmt samples (23%) demonstrated t(9;22) by karyotype. Each of these 14 samples was also RT-PCR. positive. Fourteen PB samples had poor growth and could not be karyotyped. The remaining 32 samples had adequate metaphases for analysis, but the Ph 1 was not detected. Many of the RT- PCR+/Ph 1 - cases showed weak but definite bands by RT-PCR. Overall, 85 of 113 samples (75%) were positive by RT- PCR. The M-bcr breakpoint site was found to be between exons 2 and 3 in 32 cases (38% of positive cases) and between exons 3 and 4 in the other 53 positive cases (62%). Sixteen of the 19 concurrent BM/PB paired samples had satisfactory RNA extraction for RT-PCR. Fifteen of these 16 BM/PB pairs had concordant RT-PCR results (14 pairs BM+/PB+, 1 pair BM-/PB-, 1 pair BM+/PB-) (Fig. 1). Only 5 of the 19 concurrent BM/PB pairs had adequate assessable metaphases for karyotyping. None of thesefivepairs had concordant karyotypic results. All five pairs demonstrated Phi in the BM specimen, but lacked Ph 1 in PB. None of the five BM/PB pairs analyzed by both RT-PCR and karyotyping had concordant results, as each pair was BM+/PB+ by RT-PCR, but BM+/PB- by karyotypic analysis. The 17 patients with hematologic disorders other than CML failed to show bcr/abl fusion by RT-PCR and lacked Phi by karyotyping FlG. 1. Reverse transcriptase-polymerase chain reaction (RT-PCR) autoradiograph for bcr/abl fusion. Patient specimens are run in duplicate. Lanes 1-4: Peripheral blood and bone marrow samples from RT-PCRpositive patient with exon 2 breakpoint. Lanes 5-8: Peripheral blood and bone marrow samples from RT-PCR-negative patient (post- bone marrow transplant). Lane 9: K562 cell line (exon 3 breakpoint control). Lane 10: Known positive patient (exon 2 breakpoint control). Lanes 11-15: Mgof K.562 RNA diluted in placental RNA. Lane 16: Placental RNA (negative control). Lane 17: Specimen blank. Lane 18: Exon 2 breakpoint control. DISCUSSION Reverse transcriptase-polymerase chain reaction for analysis of bcr/abl fusion has become widely available for diagnosis and follow-up of patients with CML. However, karyotyping has remained the "gold standard" for identifying the Ph 1 and may also detect additional cytogenetic abnormalities of clinical importance. In this study, RT-PCR for bcr/abl fusion was more sensitive than karyotyping in both pre-bmt and post-bmt patients. Two cases in which the Phi was "masked," not visible by karyotyping, were identified by RT-PCR. In addition, several samples with poor culture growth, including multiple PB samples, were positive by RT-PCR. The RT-PCR assay appears to have good specificity. Analysis of BM samples from patients with chronic myelomonocytic leukemia and myeloproliferative disorders other than CML did not identify bcr/abl fusion by RT-PCR or Phi by karyotyping. A.J.C.P.-June 1996

4 WELLS ET AL. 759 RT-PCR for bcr/abl in CML Reverse transcriptase-polymerase chain reaction works well on PB samples. The correlation with BM RT- PCR results is excellent (94%). Our results indicate that RT-PCR is superior to karyotyping in PB samples, offering greater sensitivity and better correlation with BM results. Reverse transcriptase-polymerase chain reaction performed on PB samples may be of clinical utility, particularly in monitoring patients with CML after therapy. Indeed, RT-PCR may obviate the need for BM aspiration and biopsy in selected patient groups (ie, patients receiving alpha-interferon who may have transient loss of Ph l 9 "' 3 or patients 6 months or more post-bmt). Interpretation of RT-PCR bcr/abl results after BMT may be problematic. Numerous conflicting reports have been generated regarding the significance of a positive RT-PCR after BMT It appears clear that many patients (up to 75%) will be RT-PCR+ for some time after BMT.' 525 Most of these will lack the Ph 1 by karyotyping. Some have suggested that persistent RT-PCR+ results for more than 6 months post-bmt may indicate increased risk for relapse. 26 Close monitoring and analysis of at least two sequential samples by RT-PCR are recommended before drawing conclusions about persistence of the bcr/abl-producing clone and presence of minimal residual disease. 24 Although the clinical significance of the positive RT-PCR after BMT remains uncertain, many of the RT-PCR+/Ph 1 - cases in this study showed weak, but definite bands by RT-PCR, suggesting a low level of the bcr/abl gene. All of these patients were studied less than 1 year after BMT and most were from 1 to 6 months post-bmt. Continued monitoring will be essential to determine the significance of these findings. Our results indicate that RT-PCR for detection of the bcr/abl fusion gene is a sensitive and specific assay. Reverse transcriptase-polymerase chain reaction is superior to karyotyping in selected settings, particularly in patients with a "masked" Phi or in PB samples, where it provides greater sensitivity and better correlation with results from BM samples. Acknowledgments. The authors thank Karen Abbott, BS, Margie Chenggis, MT (ASCP) and Cathy Thurmond, MT(ASCP) for their work in the development and performance of the RT-PCR assay. We also thank Martine Pierre-Jules for excellent secretarial assistance. REFERENCES 1. Rowley JD. A new consistent chromosomal abnormality in chronic myelogenous leukemia identified by quinacrine fluorescence and Giemsa staining. Nature 1973;243: Bartram CR, deklein A, Hagemeyer A, et al. Translocation of c- abl oncogene correlates with the presence of the Philadelphia chromosome in chronic myelocytic leukemia. Nature 1983; 306: Groffen J, Stephenson JR, Heisterkamp N, deklein A, Bartram CR, Grosveld G. Philadelphia chromosomal breakpoints are clustered within a limited region, bcr, on chromosome 22. Cell 1984;36: Kurzrock R, Gutterman JU, Talpaz M. The molecular genetics of Philadelphia chromosome-positive leukemias. N EnglJ Med 1988;319: Rowley JD. Phi-positive leukemia including chronic myelogenous leukemia. Clin Haematol 1980;9: Kawasaki ES, Clark SS, Coune MY, et al. Diagnosis of chronic myeloid and acute lymphocytic leukemias by detection of leukemia-specific mrna sequences amplified in vitro. Proc Natl AcadSci USA 1988;85: Southern LM. Detection of specific sequences among DNA fragments separated by gel electrophoresis. JMol Biol 1975;98: Mitelman F, ed. Guidelines for Cancer Cytogenetics: Supplement to An International System for Human Cytogenetic Nomenclature. Basel, Switzerland: S. Karger, Lee M, Kantarjian H, Talpaz M, et al. Detection of minimal residual disease by polymerase chain reaction in Philadelphia chromosome-positive chronic myelogenous leukemia following interferon therapy. Blood 1992; 79: Opalka B, Wandl UB, Becher R, et al. Minimal residual disease in patients with chronic myelogenous leukemia undergoing longterm treatment with recombinant interferon <v-2b alone or in combination with interferon 7. Blood 1991;78: Yoffe G, Blick M, Kantarjian H, Spitzer G, Gutterman J, Talpaz M. Molecular analysis of interferon-induced suppression of Philadelphia chromosome in patients with chronic myeloid leukemia. Blood 1987;69: Opalka B, Kloke O, Bartram CR, et al. Elimination by interferona of malignant clone in chronic myeloid leukaemia (Letter). Lancet 1989; 1: Talpaz M, Kantarjian H, Kurzrock R, Truijillo JM, Gutterman JU. Interferon-a produces sustained cytogenetic responses in chronic myelogenous leukemia. Ann Intern Med 1991; 114: Roth MS, AntiJH, Bingham EM, Ginsberg D. Detection of Philadelphia chromomose-positive cells by the polymerase chain reaction following bone marrow transplant for chronic myelogenous leukemia. Blood 1989;74: Roth MS, Antin JH, Ash R, et al. Prognostic significance of Philadelphia chromosome-positve cells detected by the polymerase chain reaction after allogeneic bone marrow transplant for chronic myelogenous leukemia. Blood 1992;79: Gabert J, Lafage M, Maraninchi D, Thuret 1, Carcassone Y, Mannoni P. Detection of residual bcr/abl translocation by polymerase chain reaction in chronic myeloid leukemia patients after bone marrow transplantation. Lancet 1989:2: Sawyers CL, Timson T, Kawasaki ES, Clark SS, Witte ON, Champlin R. Molecular relpase in chronic myelogenous leukemia patients after bone marrow transplantation detected by polymerase chain reaction. Proc Natl Acad Sci USA 1990; 87: Delfau M-H, Kerckaert J-P, d'hooghe MC, et al. Detection of minimal residual disease in chronic myeloid leukemia patients after bone marrow transplantation by polymerase chain reaction. Leukemia 1990;4: Hughes T, Martiat P, Morgan G, Sawyers C, Witte JM. Significance of residual leukaemia transcripts after bone marrow transplant for CML (Letter). Lancet 1990; 335: Hughes T, Morgan GJ, Martiat P, Goldman JM. Detection of residual leukemia after bone marrow transplant for chronic myeloid leukemia: Role of polymerase chain reaction in predicting relapse. Blood 1991;77: Martiat P, Maisin D, Philippe M, et al. Detection of residual bcr/ abl transcripts in chronic myeloid leukaemia patients in com- Vol. 105-No. 6

5 760 HEMATOPATHOLOGY plete remission using the polymerase chain reaction and nested primers. Br J Haematol 1990;75: Pignon JM, Henni T, Amselem S, et al. Frequent detection of minimal residual disease by use of the polymerase chain reaction in long-term survivors after bone marrow transplantation for chronic myeloid leukemia. Leukemia 1990;4: Lange W, Herkert R, Finke J, et al. Apparent decrease and elimination of bcr/abl mrna-expressing residual cells in patients with chronic myelogenous leukemia after allogeneic bone marrow transplantation. Ann Hematol 1991; 63: Guerrasio A, MartinelliG, SaglioG,etal. Minimal residual disease status in transplanted chronic myelogenous leukemia patients: Low incidence of polymerase chain reaction positive cases among 48 long disease-free subjects who received ummanipulated allogeneic bone marrow transplants. Leukemia 1992;6: Lee M, Khouri I, Champlin R, et al. Detection of minimal residual disease by polymerase chain reaction of bcr/abl transcripts in chronic myelogenous leukaemia following allogeneic bone marrow transplantation. Br J Haematol 1992;82: Westbrook CA. The role of molecular techniques in the clinical managment of leukemia: Lessons from the Philadelphia chromosome. Cancer 1992; 70(suppl): A.J.C.P.-June 1996