Quantitative measurement of BCR/abl transcripts using real-time polymerase chain reaction

W.-I. Lee1, H. Kantarjian2, A. Glassman3, M. Talpaz4 and M.-S. Lee1,+

1Molecular Diagnostics Laboratory, 3Cytogenetic Laboratory, Division of Pathology and Laboratory Medicine, 2Department of Leukemia, 4Department of Immunology and Biological Therapy, Division of Medicine, The University of Texas MD Anderson Cancer Center, Houston, USA

Received 2 May 2001; revised 13 November 2001; accepted 11 December 2001.


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background

Quantitative real-time polymerase chain reaction (Q-Rt-PCR) is a new tool in the detection and quantification of the BCR/abl fusion transcripts in chronic myelogenous leukemia (CML). This study investigates its specificity, sensitivity and potential clinical usefulness.

Patients and methods

Parallel analysis of Q-Rt-PCR and the conventional reverse transcription-mediated PCR (RT–PCR) were performed on 567 samples from 481 patients. Treatment response was monitored by Q-Rt-PCR at 6 and 12 months of 61 patients on STI-571 and 103 patients on interferon.

Results

The concordance rate between Q-Rt-PCR and RT–PCR was 96.3% (546/567), with 341 positives and 205 negatives. The positive equivalents ranged from 2 x 10–6 to 1.2 µg of K562 cell RNA. Karyotyping in 372 samples revealed excellent correlation with Q-Rt-PCR measurements (P <0.001). Setting residual BCR/abl <0.01 as an early goal of molecular response, we observed that STI-571 induced a better response than interferon: 49% (20 of 41 patients) versus 35% (15 of 62 patients) at 6 months (P = 0.025) and 52% (32 of 61 patients) versus 34% (35 of 103 patients) at 12 months (P = 0.01), respectively.

Conclusions

Q-Rt-PCR provides reliable measurements of BCR/abl fusion transcripts. It is potentially useful in assessing molecular residual disease after therapy.

Key words: chronic myelogenous leukemia, interferon, real-time polymerase chain reaction, STI-571


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The Philadelphia chromosome (Ph), a reciprocal translocation of the long arms of chromosomes 9 and 22, is found in >90% patients with chronic myelogenous leukemia (CML) [1] and 15–25% of patients with acute lymphoblastic leukemia (ALL) [2]. This translocation transposes the c-abl oncogene from chromosome 9q34 to the BCR gene on chromosome 22q11 [3]. The fused BCR/abl gene and its gene products provide specific markers for diagnosis and disease monitoring. Polymerase chain reaction (PCR) permits specific and sensitive detection of Ph-positive cells through the amplification of the BCR/abl fusion transcripts [4, 5]. Owing to its high sensitivity, this assay is a powerful tool for the detection of subclinical minimal residual disease; however, its clinical usefulness is hampered by difficulties in quantification.

Recently, PCR technology has evolved into real-time fluorescence detection. One of the detection methods takes advantage of the 5'->3' exonuclease activity of Taq DNA polymerase, which hydrolyzes a double-labeled fluorogenic probe upon annealing to the target PCR products during the PCR sequence extension phase [6]. Since fluorescence acquisition is monitored cycle-by-cycle, product analysis can be performed while PCR is in progress. It provides not only detection but also quantification of the accumulated target PCR products. To validate quantitative real-time PCR (Q-Rt-PCR) in the detection of BCR/abl fusion transcripts, we performed side-by-side comparison analysis with conventional reverse transcription-mediated PCR (RT–PCR). To determine the reliability of Q-Rt-PCR in quantification, the measurements of BCR/abl fusion transcripts were correlated with the results of cytogenetic analysis. To explore its potential clinical usefulness, we investigated the use of Q-Rt-PCR in assessing the molecular residual disease after therapy in Ph-positive CML patients. We were particularly interested in comparing interferon-based therapy [7] with a new treatment, STI-571 (a potent and specific tyrosine kinase inhibitor), which demonstrated promising therapeutic efficacy in recent clinical trials [810].


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient characteristics and sample procurement
A total of 567 bone marrow aspirates was procured from 470 patients and 11 normal donors who visited The University of Texas M.D. Anderson Cancer Center. These samples were obtained after informed patient consent. Of the 470 patients, 384 had CML, 35 had ALL and 51 had other hematological disease. Of the 384 CML patients, 378 were known to have the Ph chromosome by cytogenetic analysis, three were Ph negative with BCR rearrangement documented by Southern blot analysis, and three were Ph negative without BCR rearrangement and BCR/abl fusion transcripts. The clinical status of CML patients varied at the time of sample procurement: 164 in complete cytogenetic remission, 221 in chronic phase, 31 in accelerated phase, 21 in blast crisis and 21 in disease relapse during therapy. Of the 35 ALL patients, 11 had the Ph chromosome at diagnosis. The results of these patients were included in the parallel analysis with conventional RT–PCR because quantification was not the primary interest. However, these patients were excluded from the study for correlation between Q-Rt-PCR and cytogenetic analysis because our positive control was K562 cells carrying the CML type of BCR break points that might not be comparable with expression resulting from the ALL type of BCR break points. Negative controls included 11 normal donors, 24 with ALL, 14 with acute myeloid leukemia (AML), 19 with myeloproliferative diseases and 18 with other hematological diseases. All of them were shown to be Ph negative.

Q-Rt-PCR assay for the detection and quantification of the BCR/abl fusion transcripts
RNA was extracted from bone marrow samples using Trizol reagent (Gibco-BRL, Gaithersburg, MD, USA) according to the manufacturer’s instructions. The integrity of the RNAs was determined by gel electrophoresis followed by ethidium bromide staining. Samples with intact 28S and 18S RNAs were considered adequate and were subjected to reverse transcription, which was performed on 1 µg of total RNA using random hexamers and superscript II reverse transcriptase (Gibco-BRL), as recommended by the manufacturer. The resulting cDNA was subjected to a multiplex PCR to co-amplify three different types of BCR/abl fusion transcripts, b2/a2, b3/a2 and e1/a2, and an internal standard, retinoic acid receptor-alpha (RAR{alpha}), using a modification of the methods described previously [11, 12] with the primers shown in Table 1. Also added to the PCR reactions were double-labeled fluorogenic probes specific for the c-abl gene and RAR{alpha} gene at 0.2 µM and 0.05 µM, respectively (Table 1). The c-abl probe was labeled with 6-carboxy-fluorescein (6-FAM) at the 5' end and 6-carboxytetramethyl rhodamine (TAMRA) at the 3' end. The RAR{alpha} was labeled with VIC at the 5' end and TAMRA at the 3' end. Amplification of the internal standard RAR{alpha} allowed us to normalize variations in the efficiencies of reverse transcription and PCR, and to verify the integrity of RNA samples.


View this table:
[in this window]
[in a new window]
 
Table 1. Nucleotide sequences of the primers and probes used
 
PCR was performed in the ABI PRISMTM 7700 Sequence Detection System (PE Applied Biosystems, Foster City, CA, USA) for 40 cycles under the following conditions for each cycle: 95°C for 30 s, 64°C for 30 s, and 72°C for 45 s. During PCR, fluorescence acquisition was monitored cycle by cycle to measure the accumulation of the reporter dyes, 6-FAM and VIC, respectively. Based on the cycle threshold method as recommended by the manufacturer, the observed measurements of the BCR/abl fusion transcripts (BCR/ablobsv) in the tested samples were calculated against the standard curve plots using serially diluted samples prepared from K562 leukemia cells. Likewise, the observed measurements of the internal standard, RAR{alpha}obsv, were calculated against the standard curve plots using serially diluted HL60 leukemia cells. Consequently, the normalized measurements of the BCR/abl fusion transcripts in the tested samples were determined according to the following formula: BCR/ablnorm = BCR/ablobsv x 1/RAR{alpha}obsv. The rationale for the need of such normalization was that PCR efficiencies varied from sample to sample. Therefore, the actual quantities of BCR/abl transcripts in the tested samples should be adjusted accordingly. Consequently, the actual measurements could be adjusted from BCR/ablobsv by the factors that permitted the RAR{alpha}obsv measurements in the tested samples to be normalized to the level equivalent to 1 µg of RAR{alpha}obsv in the HL60 RNA control because 1 µg RNA was used in all samples. For conventional RT–PCR analysis, after the completion of PCR cycles, 20 µl of PCR products were subjected to gel electrophoresis, Southern transfer, probe hybridization and signal detection by a chemiluminescence kit (Amersham, Piscataway, NJ, USA).

Cytogenetic analysis
Cytogenetic analysis was performed according to established procedures [13] on bone marrow samples. Karyotypes were read according to the guidelines of the International System for Human Cytogenetic Nomenclature. For most samples, 20 metaphases were examined. In some instances, 15–30 metaphases were counted. Samples with <15 metaphases were considered insufficient yields and excluded from correlation with Q-Rt-PCR measurements.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Q-Rt-PCR in the detection and quantification of BCR/abl fusion transcripts
Monitoring of the fluorescence acquisition cycle by cycle, for each tested sample, Q-Rt-PCR plotted an amplification curve of fluorescence accumulation emitted by the reporter dye (6-FAM) upon the annealing and hydrolysis of the dual-labeled c-abl probe. Shown in Figure 1A are amplification plots of the BCR/abl fusion transcripts detected from serially diluted samples of Ph-positive K562 cells and a Ph-positive CML sample (labeled H2). The 10-fold serial dilutions of K562 RNAs are on the x axis, ranging from 0.1 µg down to 10–4 µg. As the dilution factor increases, the cycle threshold (the cycle number needed to reach a preset threshold) increases and the height of the amplification plot plateau becomes lower. For example, the positive fluorescence in the 10–4 dilution surpasses the threshold at PCR cycle 32–33, and then forms an amplification curve that is clearly discriminated from the negative control. Although the curve of 10–5 is not shown, a small amplification plot above the threshold line is formed after 37–38 PCR cycles. Shown in Figure 1B are amplification plots of the internal standards, RAR{alpha}, obtained from serially diluted HL60 cell RNAs (ranging from 0.1 µg down to 10–4 µg) and the Ph-positive sample (H2) mentioned above. Based on the cycle threshold method, the observed BCR/abl fusion transcript (BCR/ablobsv) in sample H2 was 4.8 x 10–2 and the observed internal standard (RAR{alpha}obsv) was 3.9 x 10–1; therefore the normalized measurement of the BCR/abl fusion transcripts (BCR/ablnorm) would be 1.23 x 10–1.



View larger version (136K):
[in this window]
[in a new window]
 
Figure 1. Q-Rt-PCR amplification plots of the BCR/abl fusion transcripts and the internal control RAR{alpha} transcripts. (A) The BCR/abl amplification curves are plotted from serially diluted K562 RNAs, ranging from 10–1 to 10–4 µg. The BCR/abl amplification plot of a Ph-positive CML sample (labeled H2) is also shown. (B) The RAR{alpha} amplification curves are obtained from serial dilutions of HL60 RNAs ranging from 10–1 to 10–4 µg and the Ph-positive CML sample, H2, which is also shown in (A).

 
Q-Rt-PCR versus conventional RT–PCR in the detection of BCR/abl fusion transcripts
Parallel analysis of Q-Rt-PCR and conventional RT–PCR was performed on 557 samples (Table 2). Of these 567 samples, 341 were positive and 205 were negative as determined by both methods. The overall concordance rate was 96.3% (546/557). There were 21 discrepancies between the RT–PCR and Q-Rt-PCR assays. In nine cases, no abnormalities were detected by the conventional RT–PCR, but Q-Rt-PCR yielded a positive BCR/abl amplification plot (Table 3). Of these nine, six showed a unique early take-off curve with a low plateau (Figure 2). None of these had a Ph chromosome detected by cytogenetic analysis. The clinical states and diagnosis of these samples were as follows: two CML in cytogenetic remission, one ALL, one AML, one lymphoma and one multiple myeloma. Therefore, these early take-off, low plateau curves appeared to be false positives. The remaining three cases had weak positives, with the BCR/ablnorm measured as 2 x 10–4, 1.1 x 10–4 and 2.6 x 10–5. The reasons for these being negative according to conventional RT–PCR could be poor Southern transfer, suboptimal probe hybridization or signal enhancement by chemiluminescence. On the other hand, 12 weak positives shown by conventional RT–PCR were not detected by Q-Rt-PCR (Table 3, cases 10–21). Of these 12, seven had a small amplification plot that appeared under the threshold line, and five had a negative without formation of an amplification plot. Among these 12, 11 were in cytogenetic remission and one had 25% Ph chromosome after interferon therapy. These findings indicate that, occasionally, weak positives as shown by conventional RT–PCR might not be detected by Q-Rt-PCR. The sensitivity of conventional PCR was increased by as much as 100-fold owing to the use of probe hybridization and chemiluminescent signal enhancement, as opposed to analysis relying solely on gel electrophoresis and ethidium bromide staining.


View this table:
[in this window]
[in a new window]
 
Table 2. Conventional RT–PCR versus Q-Rt-PCR in the detection of the BCR/abl transcripts
 

View this table:
[in this window]
[in a new window]
 
Table 3. Characteristics of the discrepancies between conventional RT–PCR and Q-Rt-PCR
 


View larger version (37K):
[in this window]
[in a new window]
 
Figure 2. An example of false-positive BCR/abl in Q-Rt-PCR. The amplification plot of a false positive (labeled 30292) shows an early take-off curve with a low plateau. This is a sample of multiple myeloma (Table 3, sample 6). The negativity is proved by conventional RT–PCR.

 
Correlation of the Q-Rt-PCR quantification with cytogenetic analysis
To determine the reliability of Q-Rt-PCR in measuring quantities of BCR/ABL transcripts, the results of 372 CML samples were correlated with the results of cytogenetic analysis. Cytogenetic analysis of the 372 samples revealed that 139 had 100–91% Ph chromosome, 56 had 90–35%, 39 had 34–1% and 138 had 0%. The ranges of BCR/ablnorm in Q-Rt-PCR positive samples were equivalent to 1.2 x 10–6 to 1 µg of K562 cell RNA. The correlation, as shown in Table 4, revealed that high quantities of BCR/abl transcripts were associated with the presence of high percentages of Ph chromosome (P <0.001), indicating that Q-Rt-PCR provided reliable quantification of BCR/abl fusion transcripts.


View this table:
[in this window]
[in a new window]
 
Table 4. Real-time PCR quantification of the BCR/abl transcripts in Ph-positive CML: correlation with cytogenetic analysis
 
Since Ph chromosome <35% after therapy is defined as a major cytogenetic response, it raises an interesting and important question regarding how it is correlated with the Q-Rt-PCR measurements of BCR/abl fusion transcripts. In the 195 samples containing >=35% Ph chromosome, 150 (77%) had a BCR/ablnorm measurement of >=1 x 10–2. On the other hand, in the 177 samples with <35% Ph chromosome, 156 (88%) had BCR/abl fusion transcripts measurements of <1 x 10–2.

Assessing molecular response by Q-Rt-PCR after interferon and STI treatments
To investigate the potential clinical usefulness of Q-Rt-PCR, we studied 164 Ph-positive CML patients who had been treated with interferon-based therapy (103 patients) or STI-571 (61 patients). The patients’ clinical states at the time treatment started varied. Patients receiving interferon were in the chronic phase of disease. In contrast, patients treated with STI-571 were either in the accelerated phase of CML or had cytogenetic abnormalities, suggesting clonal evolution. Arbitrarily defining residual BCR/ablnorm <0.01 as an early goal of molecular response after therapy, we studied the molecular response rates within the first year of treatment (Table 5). At 6 months, 20 of 41 patients (49%) on STI-571 had residual BCR/abl fusion transcripts <0.01, as compared with 15 of 62 patients (35%) on interferon (P = 0.025). At 12 months, 32 of 61 patients (52%) on STI-571 and 35 of 103 patients (34%) on interferon had residual BCR/abl fusion transcripts <0.01 (P = 0.01).


View this table:
[in this window]
[in a new window]
 
Table 5. Molecular responsiveness within the first year of therapy
 

    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Owing to its specificity and high sensitivity, RT–PCR is one of the most powerful tools in monitoring minimal residual disease in Ph-positive CML [1416]. Without the need for post-PCR manipulation, Q-Rt-PCR is superior to conventional RT–PCR in its simplicity and cost effectiveness [1719]. More importantly, its quantification capability confers un-precedented advantages over conventional PCR. Nonetheless, Q-Rt-PCR is a relative new technology that awaits systematic investigation to validate its reliability and explore its clinical usefulness.

Employing both conventional RT–PCR and Q-Rt-PCR, we performed parallel analysis of a large number of CML samples in various clinical states alongside 11 samples obtained from normal donors as negative controls. We observed a high concordance rate of 96.3%, indicating that the sensitivity and specificity of both assays are equivalent. Although there were rare discrepancies (3.7%) in this study, parallel comparison helped to investigate the causes and the patterns of false positives and false negatives in both assays. Particularly noteworthy was the characteristic false-positive pattern observed in Q-Rt-PCR: an early take-off amplification curve with a low plateau. This could be due to autohydrolysis of the fluorogenic probe as a result of repeated heating and cooling during PCR cycles, or other unknown mechanisms. These false-positive measurements were frequently high according to Q-Rt-PCR, but cytogenetic analysis and conventional RT–PCR were negative.

Three samples showed a weak Q-Rt-PCR measurement at 2 x 10–4, 1.1 x 10–4 and 2.6 x 10–5, respectively. However, conventional RT–PCR did not detect the abnormalities. These false negativities in conventional RT–PCR were most likely due to suboptimal Southern transfer, probe hybridization or signal enhancement by chemiluminescence, because the repeated analyses by both Q-Rt-PCR and conventional RT–PCR confirmed the positive results.

There were 12 weak positives shown by conventional RT–PCR, but Q-Rt-PCR showed a small curve under the threshold level in seven samples and did not forming an amplification plot in the remaining five. In fact, 11 of these 12 false negatives in Q-Rt-PCR were in cytogenetic remission, where the number of BCR/abl fusion transcripts was very low.

To test the reliability of the Q-Rt-PCR assay in the quantification of the BCR/abl fusion transcripts, we also performed a parallel correlation test between Q-Rt-PCR and cytogenetic analysis. Although these two assays were independent, we observed a highly significant correlation in which large quantities of BCR/abl fusion transcripts were associated with a high percentage of Ph-positive cells in the samples (P <0.001). This observation indicates that Q-Rt-PCR provides reliable BCR/abl fusion transcript measurements in Ph-positive CML. However, there were a few exceptions. Large quantities of BCR/abl fusion transcripts, measuring >0.01, were observed in three samples with 0% Ph chromosome. Also, small quantities of BCR/abl fusion transcripts measuring <0.001 were observed in 13 samples with >91% Ph chromosome (Table 4). These discrepancies prompted us to review the clinical states of these patients. The first patient who had 0% Ph chromosome and a large quantity of BCR/abl transcripts (1.2 x 10–1) actually had blast crisis of CML, for which chemotherapy with HyperCVAD (hyper-fractionated combination of Cytoxan, Vincristin, Adriamycin and Dexamethasone) was given. The sample was obtained 1 month after the first cycle of chemotherapy. The second patient who showed 0% Ph chromosome and high BCR/abl transcripts (1.6 x 10–2) had disease progression from late chronic phase to accelerated phase, as evidenced by increased blasts. However, the third patient (BCR/ablnorm 1.0 x 10–2) was still in cytogenetic and clinical remission. We also reviewed the clinical history of the 13 patients who had >91% Ph chromosome with small amounts (<1 x 10–3) of BCR/abl fusion transcript. Six of these patients were on chemotherapy for accelerated phase or blast crisis. The remaining seven patients were on treatment for chronic phase CML. Although the biological and clinical significance of these unusually low or high expressers is still unclear, they are rare exceptions and do not affect the analysis of this study.

The excellent correlation between Q-Rt-PCR measurements and the results of cytogenetic analysis justified the use of Q-Rt-PCR determinations to define molecular responses after therapy. In this study, we were particularly interested in patients after interferon-based therapy and patients on STI-571 treatment. Patient results following allogeneic bone marrow/peripheral blood stem cell transplants were not presented; they will be analyzed in a separate study.

In a previous study at The University of Texas MD Anderson Cancer Center, Kantarjian et al. [20] reported that the major cytogenetic response rate of early chronic phase CML to a daily dose of interferon plus low dose ara-C combination was 50%, with a median time to achievement of 7 months. In our current study, most patients on IFN therapy were also in early chronic phase, but they received more intensive therapy, a combination of IFN-{alpha}, ara-C and homoharrintonin. However, only 34% of our patients could attain a molecular response of <0.01 at 12 months. Therefore, the question is raised of how a molecular response of <0.01 compares with a major cytogenetic response of <35% of Ph chromosome. In this study, of the 195 samples containing >=35% of Ph chromosome, 150 (77%) had BCR/abl fusion transcripts measured as >=1 x 10–2. On the other hand, in the 177 samples with <35%, 156 (88%) had BCR/ablnorm measured as <1 x 10–2. Although it is still too early to draw a conclusion, these findings support the use of residual BCR/abl fusion transcripts <0.01 as the early goal of molecular response within the first year of interferon or STI-571 treatment.

Based on this criteria, we observed that the response rate of STI-571 treatment was better than interferon therapy: 49% versus 34% at 6 months and 52% versus 34% at 12 months (P = 0.025 and 0.01, respectively). Of note, patients receiving STI-571 treatment were in accelerated phase or clonal evolution as opposed to most patients on interferon therapy being in early chronic phase. The better molecular response rates observed in STI-571 treatment suggested the potential clinical usefulness of STI-571. Nonetheless, long-term clinical follow-up is needed to determine whether the early molecular response will eventually translate into durable therapeutic effectiveness.

Since Q-Rt-PCR had a much higher sensitivity than cytogenetic analysis, Q-Rt-PCR helped identify a subset of complete cytogenetic responders whose leukemia cell burden was extremely low or undetectable. Moreover, in those patients with Ph-positive cells detected by cytogenetic analysis, the Q-Rt-PCR measurements of the BCR/abl fusion transcripts correlated very well with the amounts of Ph-positive cells as determined by cytogenetic analysis. Therefore, Q-Rt-PCR reliably measures leukemia cell burden, ranging from full-blown disease to extremely small numbers and complete freedom of Ph-positive cells. More importantly, the quantitative assessment could provide unprecedented valuable information that could help predict treatment response, drug resistance and imminent disease recurrence. If the clinical significance and prognostic values of the Q-Rt-PCR assay are confirmed by other studies and multi-center collaboration, it will eventually become a gold standard for disease monitoring in Ph-positive CML.


    Acknowledgements
 
This work was supported in part by the Leukemia and Lymphoma Society of America and the Su-Ling and Shirley Lee Fund.


    Footnotes
 
+ Correspondence to: Dr M.-S. Lee, Molecular Diagnostics Laboratory, Box 54, Division of Pathology and Laboratory Medicine, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA. Tel: +1-713-792-5059; Fax: +1-713-792-4840; E-mail: mslee@mdanderson.org Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Rowley JD. A new consistent chromosome abnormality in chronic myelogenous leukemia. Nature 1973; 243: 290–293.[ISI][Medline]

2. LeBeau MM. Cytogenetic analysis of hematologic malignant diseases. In Barch MJ (ed): The ACT Cytogenetics Laboratory Manual, 2nd edition. New York, Raven Press 1991; 395.

3. Shtivelman E, Lifshitz B, Gale RP et al. Fused transcript of abl and bcr genes in chronic myelogenous leukemia. Nature 1985; 315: 550–554.[ISI][Medline]

4. Kawasaki ES, Clark SS, Coyne MY et al. Diagnosis of chronic myelogenous leukemia and acute leukemia by detection of leukemia-specific mRNA sequences amplified in vitro. Proc Natl Acad Sci USA 1988; 85: 5689–5692.

5. Lee M-S, Lemaistre A, Kantarjian HM et al. Detection of two alternative bcr/abl mRNA junctions and minimal residual disease in Philadelphia chromosome positive chronic myelogenous leukemia by polymerase chain reaction. Blood 1989; 73: 893–897.

6. Holland PM, Abramson RD, Watson R et al. Detection of specific polymerase chain reaction product by utilizing the 5' to 3' exonuclease activity of thermus acquaticus DNA polymerase. Proc Natl Acad Sci USA 1991; 88: 7276–7280.[Abstract]

7. Faderl S, Talpaz M, Estrov Z et al. Chronic myelogenous leukemia: biology and therapy. Ann Int Med 1999; 131: 207–219.[Abstract/Free Full Text]

8. Druker BJ, Sawyers CL, Talpaz M et al. Phase I trial of a specific abl tyrosine kinase inhibitor, CGP 57148, in interferon refractory chronic myelogenous leukemia patients. Proc Annu Meet Am Soc Clin Oncol 1999; 18: 24a.

9. Talpaz M, Silver RT, Druker B et al. A phase II study of STI 571 in adult patients with Philadelphia chromosome positive chronic myeloid leukemia in accelerated phase. Blood 2000; 96: 469a.

10. Kantarjian H, Sawyers C, Hockhaus A et al. Phase II study of STI571, a tyrosine kinase inhibitor, in patients with resistant or refractory Philadelphia chromosome-positive chronic myeloid leukemia. Blood 2000; 96: 470a.

11. Lee M-S, Khouri I, Champlin R et al. Detection of minimal residual disease by polymerase chain reaction of bcr/abl transcripts in chronic myelogenous leukemia following allogeneic bone marrow transplantation. Br J Haematol 1992; 82: 708–714.[ISI][Medline]

12. Cross NC, Melo JP, Feng L et al. An optimized multiplex polymerase chain reaction (PCR) for detection of BCR–ABL fusion mRNAs in hematologic disorders. Leukemia 1994; 8: 186–189.[ISI][Medline]

13. Sandberg AA, Bridge JA. Techniques in cancer cytogenetics: an overview and update. Cancer Invest 1992; 10: 163.

14. Faderl S, Talpaz M, Kantarjian HM et al. Should polymerase chain reaction analysis to detect minimal residual disease in patients with chronic myelogenous leukemia be used in clinical decision making? Blood 1999; 93: 2755–2759.[Free Full Text]

15. Hochhaus A, Reiter A, Saussele S et al. Molecular heterogeneity in complete cytogenetic responders after interferon-alpha therapy for chronic myelogenous leukemia: low level of minimal residual disease are associated with continuing remission. German CML study group and the UK MRC CML study group. Blood 2000; 95: 62–66.[Abstract/Free Full Text]

16. Chomel JC, Brizard F, Veistein A et al. Persistence of bcr-abl genomic rearrangement in chronic myeloid leukemia patients in complete and sustained cytogenetic remission after interferon-alpha therapy or allogeneic bone marrow transplantation. Blood 2000; 95: 404–408.[Abstract/Free Full Text]

17. Elmaagacli AH, Beelen DW, Opalka B et al. The amount of bcr–abl fusion transcripts detected by the real-time quantitative polymerase chain reaction method in patients with Philadelphia chromosome positive chronic myeloid leukemia correlates with the disease stage. Annal Hematol 2000; 79: 424–431.[ISI][Medline]

18. Emig M, Saußele S, Wittor H et al. Accurate and rapid analysis of residual disease in patients with CML using specific fluorescent hybridization probes for real time quantitative RT–PCR. Leukemia 1999; 13: 1825–1832.[ISI][Medline]

19. Hochhaus A, Weisser A, Rosee PL et al. Detection and quantification of residual disease in chronic myelogenous leukemia. Leukemia 2000; 14: 998–1005.[ISI][Medline]

20. Kantarjian HM, O’Brien S, Smith TL et al. Treatment of Philadelphia chromosome positive early chronic myelogenous leukemia with daily doses of interferon alpha and low-dose cytarabine. J Clin Oncol 1999; 17: 284–292.[Abstract/Free Full Text]