We monitored BCR–ABL transcript levels by quantitative real-time PCR in 103 patients treated with imatinib for chronic myeloid leukaemia in chronic phase for a median of 30.3 months (range 5.5–49.9) after they achieved complete cytogenetic remission (CCyR). The patients could be divided into three groups: (1) in 32 patients transcript levels continued to decline during the period of observation (nadir BCR–ABL/ABL ratio 0.015%); in five of these patients BCR–ABL transcripts became undetectable on repeated testing, (2) in 42 patients the transcript levels reached a plateau and (3) in 26 patients transcript numbers increased and the initial CCyR was lost. Three patients were not evaluable. Patients who remained in CCyR for at least 24 months appeared to have a low risk of subsequent cytogenetic relapse. We conclude that the pattern of ‘residual’ disease after achieving CCyR on imatinib is variable: some patients in CCyR show a progressive reduction in the level of residual disease, some reach a plateau where transcript numbers are relatively stable and others relapse with Ph-positive metaphases.
The defining feature of chronic myeloid leukaemia (CML) is the BCR–ABL fusion gene usually associated with a Philadelphia (Ph) chromosome in the patient's leukaemia cells. Until recently, interferon-alfa was the only nontransplant approach able to reduce the proportion of Ph-positive metaphases in a patient's bone marrow, and by implication the overall quantity of leukaemia cells in a patient's body. The introduction of imatinib mesylate is clearly a major advance in the management of patients with CML in chronic phase; it induces complete cytogenetic responses (CCyR) in approximately 40% of patients treated after failure of interferon-alfa1 and at least 70% of patients who start treatment soon after diagnosis.2 In comparison with earlier treatments, imatinib appears to prolong survival or progression-free survival in both categories of patients and the clinical benefit may be greatest in patients who achieve the greatest reduction in the leukaemia cell burden.1, 2, 3
Once patients have reached CCyR, their levels of BCR–ABL transcripts can be monitored by the quantitative reverse transcriptase polymerase chain reaction (Q-PCR); results are usually expressed as a percentage ratio related to an internal control transcript. The various patterns of evolution in transcript levels after CCyR is achieved are not yet fully defined in individual patients who continue on treatment with imatinib. Here we report the results of serial transcript measurements in 103 patients who achieved CCyR on treatment with imatinib for CML in chronic phase.
Between January 2000 and April 2004, 246 adult patients with BCR–ABL-positive CML in chronic phase were treated with imatinib at the Hammersmith Hospital in London. Of the 246 patients, 103 (42%) achieved CCyR and were analysed further in this study. Of these 103 patients, 69 had been treated previously with interferon-alfa and 34 were newly diagnosed (Table 1). Chronic phase was defined by conventional criteria.4 The Sokal and Euro prognostic scores were calculated as described.5, 6 Imatinib was administered as described by others;1, 7 briefly, patients received imatinib at a daily oral dose of 400 mg; no concomitant chemotherapy was administered other than initial short courses of hydroxyurea or anagrelide when deemed necessary. Imatinib dosage was adjusted depending on tolerance and response; doses were reduced in the presence of grade III–IV thrombocytopenia or neutropenia, but the aim was always to maintain the dose at or above 300 mg/day. Patients were censored if the dose had to be increased above 400 mg/day or if other antileukaemia drugs were added to the imatinib. The frequency of bone marrow morphology and cytogenetic assessments varied according to the clinical protocol used to treat patients.1, 7 In general, bone marrow examinations were undertaken at 3-month intervals for the first 2 years (or for 1 year in newly diagnosed patients); thereafter, marrow examinations were performed less frequently unless Q-PCR values suggested the possibility of a cytogenetic relapse. Cytogenetic studies were performed with standard methods using Giemsa banding. CCyR was defined by the failure to detect any Ph-positive metaphases in a bone marrow examination with a minimum of 30 metaphases examined, and was confirmed by a subsequent cytogenetic analysis. Cytogenetic relapse (loss of CCyR) was defined by the detection of one or more Ph-positive marrow metaphases, again confirmed by a subsequent cytogenetic study. Transcript numbers were measured by Q-PCR as soon as possible after CCyR was achieved and at 4- to 6-week intervals thereafter, irrespective of the protocol on which the patients were being treated.
Quantitation of BCR–ABL transcripts
We routinely collected 10–20 ml of peripheral blood for Q-PCR studies. In brief, the red cells were lysed and residual cells were homogenised in 1 ml of guanidinium thiocyanate containing β-mercaptoethanol until the solution was nonviscous as previously described.8 Total RNA was extracted from 350 μl of GTC cell lysate using RNeasy Mini kits (Qiagen, Crawley, UK) according to the manufacturer's instructions. The total RNA was reverse transcribed to cDNA using standard molecular biology techniques8 and BCR–ABL and ABL transcripts were quantified. This was achieved by subjecting the synthesised cDNA to 50 cycles of Q-PCR using the ABI 7700 Sequence Detection System (Appliedbiosystems, Foster City, CA, USA) and TaqMan Universal Master Mix in accordance with the manufacturer's instructions in a final reaction volume of 25 μl.9 Probes and primers were designed using the Primer Express software (Appliedbiosystems) to detect e13a2 (b2a2) and e14a2 (b3a2) junctions in a single reaction by Q-PCR. The BCR–ABL 6-carboxyfluorescein IndexTerm(FAM)-cccttcagcggccagtagcatctga-6-carboxy-tetramethyl-rhodamine (TAMRA) and ABL (IndexTermFAM-tgcttctgatggcaagctctacgtctcct-TAMRA) probes were dual labelled with FAM and TAMRA. The primers used for Q-PCR were: BCR–ABL forward primer: IndexTerm5′-tccgctgaccatcaayaagga-3′; BCR–ABL reverse primer: IndexTerm5′-cactcagaccctgaggctcaa-3′; ABL forward primer: IndexTerm5′-gatacgaagggagggtgtacca-3′: ABL reverse primer: IndexTerm5′ctcggccagggtgttgaa-3′. The degenerate Y base was included in the forward BCR–ABL primer to allow for the single-nucleotide polymorphism in exon 13 of the BCR gene. The BCR–ABL probes and primers were designed in association with the Europe Against Cancer (EAC) collaborative study.10 The probe and primer concentrations for ABL mRNA quantification were 200 and 300 nM, respectively, with 3 μl of cDNA. The BCR–ABL mRNA levels were measured using 100 nM of probe and 300 nM of each primer with 5 μl of cDNA. The BCR–ABL and ABL copy numbers were calculated by comparison with the standard curve generated using serial dilutions of linearised pNC210/G plasmid, containing the BCR–ABL insert described previously.8 The results of quantifying BCR–ABL transcripts were expressed as percentage ratios relative to total ABL transcripts. In our study, patients who achieved CCyR almost always had BCR–ABL/ABL ratios below 2%. Patients with BCR–ABL/ABL percentage ratios below 0.0001% were regarded as having achieved complete molecular responses. Samples with an ABL control <1 × 104 were considered suboptimal and were excluded from the analysis. The sensitivity of BCR–ABL detection for individual samples was approximated as the number of ABL transcripts found in the same volume of cDNA. The failure to detect any BCR–ABL transcripts using Q-PCR therefore represents a 4 log or greater reduction from baseline. The sensitivity in our laboratory of quantitative Q-PCR is similar to that of conventional nested PCR.9 The median BCR–ABL/ABL ratio in untreated patients in this series was 75.1% (range 42–100%, n=103). In the cases where BCR–ABL transcripts were undetectable by Q-PCR, the results were confirmed by nested primer PCR as previously described.8 The BCR–ABL was reported as undetectable if no BCR–ABL transcripts were found and the control ABL value was ⩾1 × 104. A patient was reported to have achieved complete molecular remission if BCR–ABL transcripts were undetectable in three consecutive samples tested at least 4 weeks apart. Samples from patients without detectable BCR–ABL transcripts had ABL control levels comparable to those found in samples in which the BCR–ABL transcripts were still detectable.
The probability of cytogenetic relapse was calculated using the method of cumulative incidence, where cytogenetic relapse was the event of interest and death the competitor. Univariate analyses to identify prognostic factors for cytogenetic relapse were carried out using the log-rank test. Variables found to be significant at the P<0.25 level were entered into a proportional hazards regression analysis, and a forward stepping procedure was employed to find the best model.
We classified patients according to the evolution of transcript numbers after achieving CCyR (see below). We first identified a cohort who lost their CCyR and excluded them from further analysis; we then divided the remaining patients according to whether their transcript numbers appeared to reach a ‘plateau’ or continued a downward trend during the period of observation. For this analysis, a Q-PCR result that differed from the immediately preceding value by more than 0.5 on a log10 scale was ignored unless it was confirmed by a subsequent sample or unless contemporaneous cytogenetic studies showed cytogenetic relapse. Where the ‘discordant’ transcript value was the most recent value, patients were censored from this analysis at the time of the immediately preceding transcript measurement. This occurred on three occasions. Plateau and non-plateau groups were compared using a t-test for continuous variables and a Fisher's exact test for qualitative variables. The assumption of normal distribution was verified using the Shapiro–Wilk test. P-values were two-sided and confidence intervals (CI) refer to 95% boundaries.
The median followup after achieving CCyR was 30.3 months (range 5.5–49.9 months). Of the 103 patients, 81 (80%) were followed for more than 24 months after CCyR. In all, 77 (75%) patients remained in continuous CCyR during the period of their follow-up. In 17 (16.5%) patients the BCR–ABL fusion transcript was not detectable by quantitative real-time and nested primer PCR on at least one occasion during followup; in five (4.8%) of these 17 patients BCR–ABL transcripts studied serially continued to be undetectable including the most recent study for periods of 9.6+, 11.0+, 13.1+, 20.8+ and 25.2+ months respectively. The median time to achievement of complete molecular remission was 12.7 months (range 3.4–25 months).
The median time from start of imatinib therapy to achievement of CCyR was 5.8 months (range 0.9–18.9 months). In total, 85% percent of those who achieved CCyR did so within 12 months of starting treatment. At the time of first detection of CCyR, the median BCR–ABL/ABL ratio was 0.38% (CI 0.21–0.51%; range 0–6.01%).
Of the 103 patients, 26 (25.2%) progressed to cytogenetic relapse. The cumulative incidence of cytogenetic relapse at 48 months after achieving CCyR was 26.4% (CI 18.4–36.3%) (Figure 1). The median time to loss of CCyR was 12.6 months (range 5.5–31.5 months). However, for patients who remained in CCyR for at least 24 months, the cumulative incidence of loss of CCyR in the subsequent 24 months was only 2.5% (CI 0.4–13.9%). The median followup after the 24-month point was 8.2 months (range 0.2–25.9).
For the 26 patients who relapsed, the median value for the lowest BCR–ABL/ABL ratio achieved during the followup was 0.12%. None of the patients who lost their CCyR regained it without either an increase in the imatinib dose or the addition of a second agent. The median percentage of Ph-positive metaphases in the first bone marrow examination after CCyR that showed Ph positivity was 17% (range 3–100); in five patients the first positive cytogenetic result showed more than 80% Ph-positive metaphases. During subsequent followup, seven patients proceeded to 100% Ph-positivity, 12 patients became predominantly Ph-positive and seven remained Ph-positive in the range of 1–35% Ph-positive marrow metaphases.
Four (15.4%) of the patients who relapsed progressed to accelerated or blastic phases; one of the four had no preceding chronic phase identified.11 One of these patients had been classified by Sokal score as low risk, two as intermediate risk and one as high risk. Two had been previously treated with interferon-alfa and two had received imatinib as first-line therapy. These four patients had been in CCyR for 11.2, 13.1, 15.6 and 17.0 months, respectively and their lowest BCR–ABL/ABL ratios while treated with imatinib had been 0.3, 0.12, 0.09 and 0.0075%, respectively. Two of the four patients died.
In univariate analysis, none of the variables listed in Table 2 predicted for cytogenetic relapse. Similarly, none of the variables entered into multivariate analysis predicted for cytogenetic relapse.
Subclassification of patients who maintained their CCyR
Of the 77 patients who remained in CCyR during the period of observation, 74 had at least three Q-PCR studies (median 8, range 3–18) after 18 months from the onset of imatinib therapy and were therefore evaluable over reasonably long periods of time. The mean transcript levels for these 74 patients continued to decline during subsequent follow up (Figure 2). The mean of the lowest BCR–ABL/ABL ratios achieved during the followup was 0.04% (CI 0.03–0.075%); the median time from start of imatinib to the lowest value was 10.5 months (CI 17.1–26.3). At the time of latest followup, the mean transcript level was 0.12% (CI 0.075–0.19%).
It proved possible to divide the 74 patients into two groups on the basis of serial measurements of transcript levels. In one group, the transcript levels continued to decline until the most recent value or until they fell below the level of detection. In the other group, the transcript levels reached a ‘plateau’. We defined an individual patient as having reached a plateau when the log10 of the most recent BCR–ABL/ABL ratio was no more than 0.25 ‘lower’ (equivalent to a lower number of transcripts) than the log10 of the mean BCR–ABL/ABL ratios measured after 18 months from the start of imatinib. In all, 32 (43%) patients were classified as ‘continuing to decline’ and 42 (57%) as having reached a plateau as defined above. Figure 3 shows serial transcript numbers in a representative patient from each of these two groups and the basis for classifying them together with serial transcript numbers in a patient who relapsed to Ph-positivity. Figure 4 shows the evolution of transcript levels in these two groups. The mean transcript levels in the two groups did not differ greatly when CCyR was first identified; the ‘plateau’ group had a mean BCR–ABL/ABL ratio of 0.37% (CI 0.19–0.47%), while the ‘continuing to decline’ group had a mean value of 0.19% (CI 0.075–0.47%). However, as followup continued, the mean BCR–ABL/ABL ratios for the two groups diverged, as shown clearly in Figure 4.
Four (9.5%) of the patients who met the definition of plateau had most recent transcript levels that exceeded by at least 0.25 logs their mean log levels measured during the 18 months from the start of imatinib; the actual values were 0.31, 0.31, 0.32 and 0.45 above the respective means. Repeating the analyses with exclusion of these four patients did not appreciably alter the features of the plateau group of patients described above. Longer followup will be required to determine whether these patients are in fact proceeding to cytogenetic relapse or whether their transcript levels are merely oscillating over a relatively wide range. Patients in the ‘plateau’ and ‘continuing to decline’ groups were comparable for all the variables shown in Table 1.
Only a minority of patients with CML in chronic phase treated with interferon-alfa achieve a CCyR, but such patients appear to survive significantly longer than those who fail to achieve CCyR.12 The extent to which this relatively good prognostic factor may apply also to patients responding to treatment with the new ABL kinase inhibitor imatinib cannot yet be ascertained. However, several groups have reported that patients whose leukaemia cell burden is reduced to relatively low levels as reflected by a low level of BCR–ABL transcripts in the blood appear to have a favourable progression-free survival, although the period of followup is still relatively short.1, 2
Of the 246 patients evaluable patients with CML in chronic phase treated at our institution, 103 (42%) achieved a CCyR during the period of observation. The incidence of CCyR for patients who had previously received interferon-alfa and for newly diagnosed patients was 36 and 63%, respectively (data not shown), results which are similar to those in published reports.1, 7 In this study, the cumulative incidence of cytogenetic relapse at 4 years after achieving CCyR was 26.4%; there was no significant difference between patients previously treated with interferon-alfa and patients who received imatinib as primary therapy (Table 2). Of great interest, however, was the observation that the rate of cytogenetic relapse appeared to be extremely low in patients who had remained in CCyR for 2 years or longer. This observation must, however, be interpreted with caution since the number of patients with long followup is relatively small and the duration of their followup is still relatively short.
We were able to identify three general patterns of subsequent evolution in patients who achieved CCyR. Some patients relapsed to Ph-positivity, some continued to show reduction in the level of BCR–ABL transcripts during the period of observation and some showed initial reduction but then appeared to reach a plateau with no further significant fall. Of those who showed a continuing reduction of BCR–ABL transcript numbers, five had transcripts repeatedly undetectable for a median period of 13.1 months. However, it is possible that transcripts could still have been detected in some or all of these patients by use of a more sensitive assay, such as has been employed to detect BCR–ABL transcripts in normal persons.13, 14 The recognition of a plateau is a new finding. Its existence can only be recognised if sequential results for all individual patients are plotted separately, and this may explain why it has not previously been recognised. Conversely, we do not believe that achieving a ‘plateau’ during the period of observation is necessarily an immutable status. For example, some patients who satisfy our criteria for ‘plateau’ could, if followed for longer periods, still proceed to cytogenetic relapse while other patients whose level of residual disease appears to be continuing to decline could eventually also reach a plateau, which could be just above or possibly below the level of detection in the most sensitive Q-PCR assay.
None of the possible prognostic factors studied in univariate and multivariate analysis appeared to predict for loss of cytogenetic response; in particular the classical prognostic factors for survival defined at diagnosis, respectively, by the Sokal and Euro scores had no predictive value. This could of course be readily be explained by the fact that patients in this study were already selected by virtue of their having achieved CCyR. It is clear that the actual rate of reduction in transcript numbers after initiating imatinib is an important prognostic factor15 and could possibly have helped to discriminate between patients in CCyR who reached a plateau in transcript numbers and those whose numbers continued to decline; the data required for this analysis were unfortunately not available to us.
Imatinib is able to reduce the number of leukaemia cells in a patient's body to a considerable degree. It is now generally accepted that an overall measure of leukaemia cell burden, such as may be provided indirectly by bone marrow cytogenetics and/or measurement in the blood or marrow of the number of BCR–ABL transcripts, is likely to prove a reliable indicator of disease prognosis. One should introduce a note of caution. If some leukaemia ‘stem’ cells capable of regenerating leukaemia were relatively impervious to attack by imatinib16, 17 and were at the same time vulnerable to acquisition of the ill-defined additional genetic changes that may underlie disease progression, the use of imatinib might not prolong survival in every patient. The occasional observation of patients whose disease progressed to advanced phase without any observed intervening relapse to chronic phase11 suggests that this sequence of events is possible, though hopefully rare.
Kantarjian H, Sawyers C, Hochhaus A, Guilhot F, Schiffer C, Gambacorti-Passerini C et al. Hematologic and cytogenetic responses to imatinib mesylate in chronic myelogenous leukemia. N Engl J Med 2002; 346: 645–652.
Hughes TP, Kaeda J, Branford S, Rudzki Z, Hochhaus A, Hensley ML et al. Frequency of major molecular responses to imatinib or interferon alfa plus cytarabine in newly diagnosed chronic myeloid leukemia. N Engl J Med 2003; 349: 1423–1432.
Marin D, Marktel S, Szydlo R, Klein JP, Bua M, Foot N et al. Survival of patients with chronic-phase chronic myeloid leukaemia on imatinib after failure on interferon alfa. Lancet 2003; 362: 617–619.
Kantarjian HM, Dixon D, Keating MJ, Talpaz M, Walters RS, McCredie KB et al. Characteristics of accelerated disease in chronic myelogenous leukemia. Cancer 1988; 61: 1441–1446.
Sokal JE, Cox EB, Baccarani M, Tura S, Gomez GA, Robertson JE et al. Prognostic discrimination in ‘good-risk’ chronic granulocytic leukemia. Blood 1984; 63: 789–799.
Hasford J, Pfirrmann M, Hehlmann R, Allan NC, Baccarani M, Kluin-Nelemans JC et al. A new prognostic score for survival of patients with chronic myeloid leukemia treated with interferon alfa. Writing Committee for the Collaborative CML Prognostic Factors Project Group. J Natl Cancer Inst 1998; 90: 850–858.
O’Brien SG, Guilhot F, Larson RA, Gathmann I, Baccarani M, Cervantes F et al. Imatinib compared with interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med 2003; 348: 994–1004.
Cross NC, Lin F, Chase A, Bungey J, Hughes TP, Goldman JM . Competitive polymerase chain reaction to estimate the number of BCR–ABL transcripts in chronic myeloid leukemia patients after bone marrow transplantation. Blood 1993; 82: 1929–1936.
Kaeda J, Chase A, Goldman JM . Cytogenetic and molecular monitoring of residual disease in chronic myeloid leukaemia. Acta Haematol 2002; 107: 64–75.
Gabert J, Beillard E, van der Velden VHJ, Bi W, Grimwade D, Pallisgaard N et al. Standardization and quality control studies of ‘real-time’ quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in leukemia – a Europe Against Cancer program. Leukemia 2003; 17: 2318–2357.
Avery S, Nadal E, Marin D, Olavarria E, Kaeda J, Vulliamy T et al. Lymphoid transformation in a CML patient in complete cytogenetic remission following treatment with imatinib. Leuk Res 2004; 28 (Suppl 1): 75–77.
Bonifazi F, de Vivo A, Rosti G, Guilhot F, Guilhot J, Trabacchi E et al. Chronic myeloid leukemia and interferon-alpha: a study of complete cytogenetic responders. Blood 2001; 98: 3074–3081.
Biernaux C, Loos M, Sels A, Huez G, Stryckmans P . Detection of major bcr–abl gene expression at a very low level in blood cells of some healthy individuals. Blood 1995; 86: 3118–3122.
Bose S, Deininger M, Gora TJ, Goldman JM, Melo JV . The presence of typical and atypical BCR–ABL fusion genes in leukocytes of normal individuals: biologic significance and implications for the assessment of minimal residual disease. Blood 1998; 92: 3362–3367.
Merx K, Muller MC, Kreil S, Lahaye T, Paschka P, Schoch C et al. Early reduction of BCR-ABL mRNA transcript levels predicts cytogenetic response in chronic phase CML patients treated with imatinib after failure of interferon alpha. Leukemia 2002; 16: 1579–1583.
Graham SM, Jorgensen HG, Allan E, Pearson C, Alcorn MJ, Richmond L et al. Primitive, quiescent, Philadelphia-positive stem cells from patients with chronic myeloid leukemia are insensitive to STI571 in vitro. Blood 2002; 99: 319–325.
Bhatia R, Holtz M, Niu N, Gray R, Snyder DS, Sawyers CL et al. Persistence of malignant hematopoietic progenitors in chronic myelogenous leukemia patients in complete cytogenetic remission following imatinib mesylate treatment. Blood 2003; 101: 4701–4707.
We thank the many medical and nursing staff involved in the care of the patients included in this analysis. We thank the technical staff responsible for the cytogenetic and molecular analyses. Some of the patients were treated in studies 110, 113 and 0106 with imatinib (GlivecR) provided at no cost by Novartis Pharma (Basel, Switzerland).
About this article
Cite this article
Marin, D., Kaeda, J., Szydlo, R. et al. Monitoring patients in complete cytogenetic remission after treatment of CML in chronic phase with imatinib: patterns of residual leukaemia and prognostic factors for cytogenetic relapse. Leukemia 19, 507–512 (2005) doi:10.1038/sj.leu.2403664
- BCR–ABL transcripts
Genome Medicine (2019)
Development and evaluation of armored RNA-based standards for quantification ofBCR-ABL1p210/p190fusion gene transcripts
Journal of Clinical Laboratory Analysis (2018)
Harnessing the power of proteomics for identification of oncogenic, druggable signalling pathways in cancer
Expert Opinion on Drug Discovery (2017)
Revista Brasileira de Hematologia e Hemoterapia (2017)
DNA and Cell Biology (2017)