Quantifying leukemia

Molecular diagnostics of hematologic malignancies

Hematologic malignancies, particularly lymphoma and leukemia, are a diverse group of diseases with a myriad of different presentations. Although the diagnosis of these diseases can be straightforward, there are many cases in which the diagnosis is difficult to establish with conventional methods. Molecular diagnostic testing to identify oncogenes and clonal lymphocyte populations can aid in resolving ambiguous cases, and several of these tests are routinely available for canine patients. Sensitive polymerase chain reaction-based methods are also useful for answering a variety of research questions. Here, detection of mutations of the c-kit gene in mast cell tumors, the bcr-abl fusion gene in myelogenous leukemias, and clonality of lymphocyte populations for the diagnosis and monitoring of lymphoma and leukemia are discussed.

Non-genetic heterogeneity--a mutation-independent driving force for the somatic evolution of tumours

Clonal populations of mammalian cells are inherently heterogeneous. They contain cells that display non-genetic variability resulting from gene expression noise and the fact that gene networks have multiple stable states. These stable, heritable variants within one cell type can exhibit different levels of responsiveness to environmental conditions. Hence, they could in principle serve as a temporary substrate for natural selection in the absence of mutations. We suggest that such ubiquitous but non-genetic variability can contribute to the somatic evolution of cancer cells, hence accelerating tumour progression independently of genetic mutations.

Clinical results with imatinib in chronic myeloid leukaemia

Since its introduction 5 years ago, imatinib mesylate has shown remarkable efficacy in treating patients with chronic myeloid leukaemia. Here we shall review the clinical results seen with imatinib at all stages of the disease, current views on the best way to monitor patients' responses and potential ways of predicting response to treatment. We shall also briefly cover the reasons why quiescent stem cells pose a theoretical threat to successful treatment.

Molecular surveillance of chronic myeloid leukemia patients in the imatinib era - evaluation of response and resistance

Residual disease in chronic myeloid leukemia patients may be assessed by various molecular methods. After imatinib treatment a significant proportion of patients achieve complete cytogenetic remission (CCR) and a sensitive method is necessary to monitor treatment response and to detect early signs of relapse. Reverse-transcriptase polymerase chain reaction (RT-PCR) is by far the most sensitive approach to assess residual disease in this group of patients. Qualitative PCR methods give only limited information about the residual leukemic mass. Quantitative RT-PCR (Q-PCR) assays enable to monitor the kinetics of residual BCR-ABL transcripts over time in patients with a good response to imatinib. Early Q-PCR results on imatinib treatment can help to identify individuals who are likely to have a good response. In chronic phase patients after CCR, Q-PCR may identify patients who are likely to continue with their CCR or to relapse and may help to optimize treatment for this group of patients. The definition of molecular surrogate endpoints beyond CCR for studies which are currently planned demands standardization of the nomenclature and of technologies to measure these targets.

Detection of minimal residual disease in acute myelogenous leukemia

Acute myelogenous leukemia (AML) is considered to be in complete remission when fewer than 5% of the cells in bone marrow are blasts. Nevertheless, approximately two thirds of patients relapse due to persisting leukemic blasts. The persistence of these cells, below the threshold of morphological detection, is termed minimal residual disease (MRD) and various methods are used for its detection. These methods include classical cytogenetics, fluorescence in situ hybridization, qualitative and quantitative RT-PCR and multiparametric flow cytometry. Currently, less than half of the AML patients have a specific marker detectable by RT-PCR techniques. The major specific molecular markers are involvement of the MLL gene with up to 50 different partners and partial tandem duplications, the core binding factor leukemias with AML1/ETO and CBFbeta/MYH11 rearrangements, PML/RARalpha in acute promyelocytic leukemia, internal tandem duplications and mutations of FLT3 and some other rare translocations. In addition, several other genes show abnormal expression levels in AML, including the Wilms tumor gene, the PRAME gene and Ig/TCR rearrangements. Most of these genetic abnormalities can be detected by qualitative but more importantly by quantitative RT-PCR. The kinetics of disappearance of molecular markers in AML differs between the various types of leukemias, although at least a 2 log reduction of transcript after induction chemotherapy is necessary for long-term remission in all types. Conversely, the change of PCR from negativity to positivity is highly predictive of relapse. Whereas in acute lymphoblastic leukemia, multiparametric flow cytometry is an established method for MRD detection, this is less so in AML. The reason is the absence of well-characterized leukemia-specific antigens and the existence of phenotypic changes at relapse. On the other hand, this method is convenient due to its simplicity and universal applicability. In conclusion, several methods can be used for MRD detection in AML patients; each has its pros and cons. Several issues still remain to be settled including the choice of the best method and the timing for MRD monitoring and above all the practical clinical implications of MRD in the various types of AML.

Imatinib therapy in chronic myeloid leukemia

This article briefly recaps the key clinical trials that involve imatinib and focuses on the ongoing results and implications of these studies for future research in imatinib-based therapy in chronic myeloid leukemia (CML). Imatinib by no means has replaced allografting in the treatment of CML, although clearly, there has been a radical shift in thinking in considering which patients should undergo transplant. The route to cure with minimal toxicity may involve creative use of drug and transplant technologies.

Frequency of major molecular responses to imatinib or interferon alfa plus cytarabine in newly diagnosed chronic myeloid leukemia

BACKGROUND

In a randomized trial, 1106 patients with chronic myeloid leukemia (CML) in chronic phase were assigned to imatinib or interferon alfa plus cytarabine as initial therapy. We measured levels of BCR-ABL transcripts in the blood of all patients in this trial who had a complete cytogenetic remission.

METHODS

Levels of BCR-ABL transcripts were measured by a quantitative real-time polymerase-chain-reaction assay. Results were expressed relative to the median level of BCR-ABL transcripts in the blood of 30 patients with untreated CML in chronic phase.

RESULTS

In patients who had a complete cytogenetic remission, levels of BCR-ABL transcripts after 12 months of treatment had fallen by at least 3 log in 57 percent of those in the imatinib group and 24 percent of those in the group given interferon plus cytarabine (P=0.003). On the basis of the rates of complete cytogenetic remission of 68 percent in the imatinib group and 7 percent in the group given interferon plus cytarabine at 12 months, an estimated 39 percent of all patients treated with imatinib but only 2 percent of all those given interferon plus cytarabine had a reduction in BCR-ABL transcript levels of at least 3 log (P<0.001). For patients who had a complete cytogenetic remission and a reduction in transcript levels of at least 3 log at 12 months, the probability of remaining progression-free was 100 percent at 24 months, as compared with 95 percent for such patients with a reduction of less than 3 log and 85 percent for patients who were not in complete cytogenetic remission at 12 months (P<0.001).

CONCLUSIONS

The proportion of patients with CML who had a reduction in BCR-ABL transcript levels of at least 3 log by 12 months of therapy was far greater with imatinib treatment than with treatment with interferon plus cytarabine. Patients in the imatinib group with this degree of molecular response had a negligible risk of disease progression during the subsequent 12 months.

Risk classification at the time of diagnosis differentially affects the level of residual disease in children with B-precursor acute lymphoblastic leukemia after completion of therapy

We have previously reported that children with B-precursor acute lymphoblastic leukemia (ALL) who remained in remission after successfully completing therapy had leukemia cells detectable by polymerase chain reaction (PCR) (N Engl J Med 1997;336(5):317-23). These patients were treated by an institutional protocol (P89-04) that lacked the post-remission intensification features of the contemporary Berlin-Frankfurt-Münster (BFM) based ALL protocols. In this report, we compared residual leukemia levels for patients on the P89-04 (n=15) and BFM-based Children's Cancer Group (CCG) studies (n=23) and for patients stratified according to risk group. Our goal was to establish which risk factors correlated with level of residual disease. Patients enrolled on the CCG protocols had lower levels of residual disease after completion of therapy than the P89-04 patients (P<0.019). Patients with high-risk disease also had lower levels of residual disease than patients with low risk disease (P<0.0001). Three-way analysis including time off treatment, risk group determined by features at presentation, and treatment protocol showed that risk group was the only significant independent variable (P<0.001).

Molecular monitoring of chronic myeloid leukemia

The tyrosine kinase inhibitor imatinib mesylate (Gleevec) (formerly STI571) has proven to be an effective and safe new therapy for patients with chronic myeloid leukemia (CML). It has induced short-term hematologic control in many patients with advanced-phase CML, with some patients achieving durable responses. In chronic-phase patients it induces significantly better cytogenetic responses and lower progression rates than interferon-alpha. However, relapse is a significant problem, especially for advanced-phase patients, and imatinib alone appears unlikely to be curative in any patient group. Real-time quantitative polymerase chain reaction (Q-PCR) provides an accurate, sensitive, and noninvasive measure of residual leukemia in patients on imatinib. Levels of BCR-ABL in the blood correlate strongly with the bone marrow cytogenetic results and early measurement can predict subsequent cytogenetic response. Complete molecular responses (no BCR-ABL detected by real-time Q-PCR) are rarely achieved. Sequential real-time Q-PCR studies should facilitate rational patient management and allow comparison of different imatinib-based treatment strategies. It may be possible to define levels of molecular response that predict long-term disease control. In addition, by defining patterns of response, an early indication of imatinib resistance may be detected.

Detection of the BCR-ABL gene by interphase fluorescence in situ hybridization (iFISH) in chronic myelogenous leukemia patients after hemopoietic stem cell transplantation: the feasibility of iFISH monitoring of therapeutic response in peripheral blood

The detection of the Philadelphia (Ph) translocation has been accomplished primarily by cytogenetic analysis and reverse transcriptase polymerase chain reaction (RT-PCR). RT-PCR is highly sensitive (1/10(4)-10(6)) but not quantitatively reliable and is thus unsuitable for the monitoring of Ph-positive cells during therapy. Interphase fluorescence in situ hybridization (iFISH) allows analysis of a large number of cells (> 500) in a timely and efficiently quantitative manner. We obtained 118 peripheral blood (PB) and 127 bone marrow (BM) samples from 75 adult chronic myelogenous leukemia (CML) patients undergoing stem cell transplantation. We simultaneously performed nested RT-PCR and iFISH for all samples. False-positive cells were detected in 2.48% +/- 0.93% (mean +/- SD) of PB samples and 2.75% +/- 0.83% of BM samples. The iFISH results for PB and BM ranged from 1.4% to 92.8% and 1.0% to 93.8%, respectively. Correlation analysis of iFISH results for PB versus BM samples showed a strong relation (r = .993). A significant correlation (P < .05) was also found between iFISH and first-round RT-PCR. The sensitivity of BCR-ABL iFISH was similar to that of first-round RT-PCR, and iFISH results for PB and BM were also well correlated. Thus, iFISH analysis of PB and/or BM samples may be more clinically reliable than RT-PCR in the quantitative monitoring of BCR-ABL fusion in CML after transplantation.

Use of real time leukaemia data to validate model predictions based on analyses and computer simulations

Predictions arising out of a mathematical model that describes the expansion of leukaemia from a diffusion-orientated perspective are critiqued and validated by employing available real time data. Based on agreements found between model predictions and the data, but mindful of the limitations it presents, it is concluded that the model could be used to describe the dynamics of normal and abnormal cells in leukaemia. It is suggested that further studies of the behaviour of certain normal cell types in contrast to abnormal cells during leukaemic development could engender additional insights into leukaemia and its treatment.

Recovery of normal hemopoiesis in disseminated cancer therapy--a model

The strategy of normal cell regeneration with recombinant hematopoietic growth factors during cancer chemotherapy is investigated by superimposing a treatment protocol on a simple model that describes an expanding malignant cell population that is coexisting with and inhibiting the population of normal cells. The model predictions suggest that the strategy of normal cell stimulation, possibly with growth factors, and possibly carried out within an intensive treatment framework may be a worthwhile chemotherapeutic option. Under this protocol, the model also predicts a minimum time interval for active treatment, a time to discontinue treatment, and a rest period during treatment in order to guarantee patient safety and recovery. Consequently, by relating and comparing model predictions to patient data, model simulations forecast that treatment could be shortened by 1-2 weeks if organized over or in the neighborhood of a predicted optimal time interval. Following this, it is conjectured that such an approach engendered by the model could produce outcomes that may have an edge over outcomes arising from therapeutic strategies that are executed over time frames that are relatively longer or significantly shorter than the predicted optimal time.

Comparison of a multiplex reverse transcriptase-polymerase chain reaction for BCR-ABL to fluorescence in situ hybridization, Southern blotting, and conventional cytogenetics in the monitoring of patients with Ph1-positive leukemias

A multiplex reverse transcriptase-polymerase chain reaction (RT-PCR) assay for both major forms of BCR-ABL was compared with fluorescence in situ hybridization (FISH), karyotyping, and Southern blotting for disease monitoring in 37 follow-up bone marrow samples from 32 patients with Ph1-positive leukemia. Of these 37 samples, 33 were from patients with chronic myeloid leukemia (CML) (26 post allogeneic bone marrow transplantation [AlloBMT] and seven during interferon-alpha therapy) and 4 from Ph1-positive acute lymphoblastic leukemia (ALL) patients (1 post AlloBMT and 3 post high dose chemotherapy). For the 27 samples studied after AlloBMT (26 CML and 1 Ph1-positive ALL) the time after transplantation ranged from 1 to 107 months (median 47.5 months). In 8 (22%) of the 37 samples there were discrepant results among methods. The discrepancy rates relative to other techniques were: karyotyping 17% (5 of 29), Southern blotting 18% (6 of 33), multiplex RT-PCR 8% (3 of 37), and FISH 8% (3 of 37). Therefore, the relative accuracy of each method for disease monitoring in Ph1-positive leukemia was: 83% (24 of 29) for karyotyping, 82% (27 of 33) for Southern blotting, 92% (34 of 37) for FISH, and 92% (34 of 37) for multiplex RT-PCR. This multiplex RT-PCR assay appears equivalent to FISH in terms of accuracy, simplicity, and turnaround time and both are superior to Southern blot and conventional cytogenetics in the laboratory monitoring of Ph1-positive leukemias.

Detection and quantification of residual disease in chronic myelogenous leukemia

The degree of tumor load reduction after therapy is an important prognostic factor for patients with CML. Conventional metaphase analysis has been considered to be the 'gold standard' for evaluating patient response to treatment but this technique normally requires bone marrow aspiration and is therefore invasive. The frequency of cytogenetic analyses can be considerably reduced if patients are also monitored by molecular methods, which can be performed on peripheral blood specimens. Of the various techniques available, most attention has been paid to RT-PCR for BCR-ABL mRNA since this is by far the most sensitive. Simple, non-quantitative RT-PCR analysis gives only limited information on patients after treatment. Quantitative RT-PCR assays have been developed to monitor the kinetics of residual BCR-ABL transcripts over time. Variables in the quantitative PCR assay may be controlled for by quantification of transcripts of a normal gene (eg ABL or glucose-6-phosphate dehydrogenase, G6PD) as an internal standard. After allogeneic stem cell transplantation, most patients become RT-PCR negative, often after a period of low level positivity that may persist for several months. Those patients destined to relapse are characterized by the reappearance and/or rising levels of BCR-ABL transcripts. In contrast, for patients treated with interferon-alpha (IFN) residual disease is rarely, if ever, eliminated. The actual level of minimal residual disease in complete cytogenetic responders to IFN correlates with the probability of relapse. New quantitative real time procedures promise to simplify the protocols that are currently in use, but standardization and the introduction of rigorous, internationally accepted controls are required to enable RT-PCR to become a robust and routine basis for therapeutic decisions.

Molecular heterogeneity in complete cytogenetic responders after interferon-α therapy for chronic myelogenous leukemia: low levels of minimal residual disease are associated with continuing remission

A substantial minority of patients with chronic myelogenous leukemia (CML) achieve a complete response (CR) to treatment with interferon- (IFN), defined as the disappearance of Philadelphia chromosome-positive metaphases. Currently it is unclear how long IFN treatment should be continued for such patients. We used a competitive reverse transcriptase-polymerase chain reaction (RT-PCR) to quantify levels of BCR-ABL transcripts in 297 peripheral blood specimens collected from 54 patients who had achieved CR with IFN. The median duration of observation was 1.9 years (range, 0.3-11.0 years). Total ABL transcripts were quantified as internal control and results were expressed as the ratio BCR-ABL/ABL. All 54 patients had molecular evidence of residual disease, although 3 patients were intermittently PCR negative. The median BCR-ABL/ABL ratio at the time of maximal response for each patient was 0.045% (range, 0%-3.6%). During the period of observation 14 patients relapsed, 11 cytogenetically to chronic phase disease and 3 directly to blastic phase. The median ratio of BCR-ABL/ABL at maximal response was significantly higher in patients who relapsed than in those who remained in CR (0.49% versus 0.021%,P < 0.0001). Our findings show that the level of residual disease falls with time in complete responders to IFN, but molecular evidence of disease is rarely if ever eliminated. The actual level of minimal residual disease correlates with the probability of relapse. We suggest that for patients who reach CR, IFN should be continued at least until relatively low levels of residual leukemia are achieved. (Blood. 2000;95:62-66)

Molecular heterogeneity in complete cytogenetic responders after interferon-α therapy for chronic myelogenous leukemia: low levels of minimal residual disease are associated with continuing remission

A substantial minority of patients with chronic myelogenous leukemia (CML) achieve a complete response (CR) to treatment with interferon- (IFN), defined as the disappearance of Philadelphia chromosome-positive metaphases. Currently it is unclear how long IFN treatment should be continued for such patients. We used a competitive reverse transcriptase-polymerase chain reaction (RT-PCR) to quantify levels of BCR-ABL transcripts in 297 peripheral blood specimens collected from 54 patients who had achieved CR with IFN. The median duration of observation was 1.9 years (range, 0.3-11.0 years). Total ABL transcripts were quantified as internal control and results were expressed as the ratio BCR-ABL/ABL. All 54 patients had molecular evidence of residual disease, although 3 patients were intermittently PCR negative. The median BCR-ABL/ABL ratio at the time of maximal response for each patient was 0.045% (range, 0%-3.6%). During the period of observation 14 patients relapsed, 11 cytogenetically to chronic phase disease and 3 directly to blastic phase. The median ratio of BCR-ABL/ABL at maximal response was significantly higher in patients who relapsed than in those who remained in CR (0.49% versus 0.021%,P &lt; 0.0001). Our findings show that the level of residual disease falls with time in complete responders to IFN, but molecular evidence of disease is rarely if ever eliminated. The actual level of minimal residual disease correlates with the probability of relapse. We suggest that for patients who reach CR, IFN should be continued at least until relatively low levels of residual leukemia are achieved. (Blood. 2000;95:62-66)

Persistent and recurrent blue nevi

Persistence of common melanocytic nevi has been fairly well characterized, clinically and histologically. In contrast, persistence of blue nevi has been reported infrequently. To define this entity better, nine cases of biologically persistent and clinically recurrent blue nevi are described. The persistent lesions in four cases were spindle-fascicular blue nevi; one showed senescent or "ancient" change and one had additional deep penetrating/epithelioid blue nevus features with atypical changes worrisome for malignancy. These changes included increased cellularity, cellular pleomorphism, mitotic figures, and a lymphocytic infiltrate. Three were biphasic dendritic-sclerotic/spindle-fascicular blue nevi, one of which had atypical changes. One case was a dendritic-sclerotic ("common") blue nevus. The original histology in one case was unavailable, but the recurrence was a combined blue nevus. The interval from initial biopsy to biopsy of the recurrent lesion was often longer (mean 2.7 years) for recurrent blue nevi than for recurrent common compound or intradermal melanocytic nevi. In addition, in contrast to recurrent common melanocytic nevi, the recurrence, in at least one case, extended beyond the scar of the original excision. These cases demonstrated that blue nevi of all histiotypes and combinations are capable of persistence with clinical recurrence. The persistence usually was histologically similar to the original, but in some cases was more "cellular" because, for the most part, the excisions of the persistent lesion revealed a deeper spindle-fascicular ("cellular") component not evident in the original superficial biopsy. In two cases, the original blue nevus appeared completely banal, but the persistent/recurrent lesions were histologically distinct and demonstrated atypical histologic features. Yet, follow-up (average 3.7 years) supports benign biology. Clinical recurrence is often associated with malignant transformation in blue nevus, but this series demonstrates that malignant tumor progression is not necessarily the case. In the absence of necrosis en mass, marked cytologic atypia, and frequent mitotic figures, the described atypical morphologic parameters in previously biopsied small blue nevi are probably reactive and "pseudomalignant." Awareness of this potential change may avoid diagnostic and prognostic errors.