Definition, Diagnostic Criteria, and Differential Diagnosis

CML is a clonal myeloproliferative expansion of transformed primitive hematopoietic progenitor cells involving myeloid, monocytic, erythroid, megakaryocytic, B-lymphoid, and occasionally T-lymphoid lineages.1 Since the 1960s when Nowell and Hungerford2 described the specific karyotypic abnormality—the Philadelphia (Ph) chromosome—there has been rapid progress in our understanding of pathogenesis, providing us with the means to easily diagnose and monitor the disease.

The detection of BCR-ABL, the fusion gene involving the breakpoint cluster region and the cellular oncogene, ABL, is the pathognomonic feature of most cases of CML (see comments below on "atypical" CML). Few conditions demonstrate overlapping features, the most common being a "leukemoid reaction" usually in response to severe infection. In this situation however, the presence of splenomegaly and a low leukocyte alkaline phosphatase (LAP) score suggest CML. The presence of the characteristic Ph chromosome will allow the distinction of CML from such disorders as primary proliferative polycythemia, idiopathic myelofibrosis, and primary thrombocythemia, which can occasionally have a somewhat confusing clinical presentation. The detection of BCR-ABL in a peripheral blood sample by reverse-transcriptase polymerase chain reaction (RT-PCR) techniques will give the definitive answer, but one should be aware that approximately 5 percent of CML cases are negative for the Ph chromosome.3,4 Among such Ph-negative patients there is a preponderance of males and older patients, with lower leukocyte counts and thrombocytopenia being more typical of this subgroup. In those patients who are negative for the Ph chromosome, 2.5 percent are also BCR-ABL negative and the prognosis is poorer in these patients.5 Specific mention should be made of those patients apparently with primary thrombocythemia in whom the Ph chromosome is detectable. Such patients should be considered to have CML in a less typical form and thus managed in a similar way. It is therefore implicit that patients with apparent primary thrombocytosis should be tested for the Ph translocation and/or BCR-ABL by RT-PCR.

Investigation of a Suspected Case of CML

The specifics of the investigation of a newly presenting patient with CML are detailed in Table 5—1. In the presenting history it may be helpful to elucidate certain features. These include the presence of night sweats or bone pain as they may indicate transforming disease. Symptoms suggestive of hyperviscosity such as headaches, confusion, and visual disturbances are important to identify. It may be helpful to determine exposure to potential mutagens. Allaying fears that the disease may be inherited is important, along with establishing whether the patient has any siblings and hence potential for allografting. Examination should particularly focus on retinal examination and lymph node areas, and include documenting the size of the spleen and liver.

The "gold standard" diagnostic test is the detection of the Ph chromosome and/or the BCR-ABL fusion gene. A full blood count and film review are critical in establishing the prognostic score (see later), and the number of blasts, basophils, and eosinophils should be particularly noted for use in calculating a prognostic score.6 If leukapheresis is considered appropriate, urgent human immunodeficiency virus (HIV) and hepatitis B and C serology will be required prior to the procedure. Before performing the bone marrow examination it may be worthwhile to consider a sample for one of the many research groups interested in CML. Details of morphology and cytogenetics should be recorded. Immunophenotyping will only be relevant to classify blast crisis. The trephine biopsy should be assessed for cellularity and degree of fibrosis.

Table 5-1. Investigations To Be Performed in Suspected Cases of CML


• Biochemistry screen including urate

• Blood group/save serum

• Coagulation screen

• Bone marrow aspirate for: Morphology

Cytogenetics (Fluorescent in situ hybridisation[FISH] if metaphase cultures fail) Sample for immunophenotyping (take an EDTA sample but only process if blast crisis is evident morphologically) Samples stored for research purposes if appropriate locally or for mailing to research group. (Often in 20 ml container with preservative-free heparin [PFH])

• Bone marrow trephine Assess cellularity Degree of fibrosis


• Urgent HIV, Hepatitis B and Hepatitis C serology if leukophere-

sis being considered

• CMV serology if allograft being considered

• HLA type patient and sibling if allograft being considered

• Peripheral blood leukocyte alkaline phosphatase (LAP) score and possibly peripheral blood BCR-ABL RT-PCR if marrow aspirate unavailable/inappropriate (LAP score is obsolete in many labs now).

• RT-PCR (useful to define breakpoint for future monitoring)

Consideration should be given to determining the cytomegalovirus (CMV) status of the patient, both for the initial blood product provision and for assessing potential for an allo-graft. Equally, HLA-typing of the patient and his or her siblings should be performed at an early stage to enable further planning of therapeutic strategy. The peripheral blood LAP score or even BCR-ABL RT-PCR may be helpful if the marrow aspirate is unavailable or inappropriate.

Modes of Presentation

Savage et al.7 described a series of 430 consecutive cases presenting to one center for consideration of allogeneic transplant. Although this group of patients is preselected, the data are the first to describe the presenting features of such a large group of CML patients. Their data are summarized in Table 5-2. Up to 20 percent of patients were asymptomatic and were diagnosed when a blood sample was taken for other reasons. Some cases had been diagnosed incidentally during pregnancy, while donating blood, or undergoing routine surgery.

Of those presenting with symptoms, the ten most commonly recorded are shown in Table 5-2. In retrospect the patient may be able to describe fatigue, weakness, or a sense of fullness in the left upper quadrant with early satiety after meals. Other symptoms including visual disturbance, weakness, arthralgia, cough, malaise, dizziness, nausea/vomiting, ankle edema, priapism, and mental changes occur in less than 5 percent of cases. Thrombo-

Table 5-2. Clinical Presentation of 430 Patients Referred to the Hammersmith Hospital for Consideration of Transplantation from 1981

• 80% of patients are symptomatic at time of presentation

• 20% of patients may have the diagnosis of CML made inciden tally on routine FBC

• 93% of patients present with chronic phase disease

• Thrombosis and leukostasis are rare even with very high platelet/white cell counts

• Median FBC values are as follows (range): WBC 174 (5.0 - 850.0) x109/L Hb 10.3 (4.9-16.6) g/dl

Plt 430 (17-3182) x109/L

• 19% of patients present with a WBC >350 x 109/L

• 25% of patients present with platelets >1 x 1012 /L

Ten Most Common Symptoms at Presentation

Fatigue and lethargy


Weight loss

Splenic discomfort

Abdominal mass or fullness


Bone pain




33.5% 21.3% 20.0% 18.6% 14.8% 14.6% 7.4% 6.2% 5.8% 4.5%

The Commonest Examination Findings

Spleen palpable (overall)

1-10 cm


Spleen not palpable Purpura Palpable liver

(Data from Savage7)

cytosis may occur with a count above 1 X 1012/L in 25 percent of cases, although there does not appear to be a correlation between bleeding and level of thrombocytosis. Splenomegaly and purpura were the most common physical signs at presentation, at 40 and 16 percent, respectively. The authors also described greater leukocytosis and splenomegaly in males and younger patients.

Phases of Disease

Some 93 percent of patients present in chronic phase,7 that is, there are less than 5 percent blasts in the bone marrow. It is possible to have a marginal excess of blasts at presentation but after starting treatment it becomes clear the patient is in chronic phase. The duration of chronic phase is usually between 3 and 8 years, with the median duration 4 to 5 years in most series but patients can present in blast crisis or, alternatively, not progress for 15 years (see section on interferon therapy later). The disease then inevitably progresses from this 'benign' stage to the accelerated phase and ultimately blast crisis. At this stage the disease is relatively resistant to treatment and thus rapidly fatal.

The definition of accelerated phase is vague but accepted clinical practice follows the criteria laid out in Table 5-3.8>9 Clinical features may signal the progression to accelerated phase with pyrexia, night sweats, weight loss, and increasing splenomegaly being typical occurrences. Laboratory features may also be indicative of "acceleration," with anemia, increasing basophils or eosinophils, thrombocytopenia, or increasing proportion of blasts being the most common findings. The appearance of hyposegmented neutrophils (i.e., with the Pelger-Huet anomaly) is also recognized.10 In detailed analysis of the ery-throid series it may be possible to detect a slight increase in hemoglobin F,11 although this is more typical of the CML of childhood.

The definition of blast crisis is based on the presence of more than 30 percent blasts in the peripheral blood or bone marrow, or the demonstration of extramedullary infiltration of blast cells. In two-thirds of cases the blasts are myeloid with one-third lymphoid. Caution is required as the morphological features may be misleading and immunophenotyping is therefore recommended in all cases. Lymphoid transformation carries a marginally better prognosis than myeloid, although both are usually fatal despite intensive treatment and have a median survival from diagnosis of blast crisis of only 3 to 6 months. It is difficult to distinguish those patients presenting with Ph-positive acute myelocytic leukemia (AML) or acute lymphocytic leukemia (ALL) from those with blast crisis of CML. Some clinicians hold that the presence of massive splenomegaly is more in keeping with pre-existing CML whereas others believe the

Table 5-3. The Criteria Used in Defining Phases of Disease in CML

Criteria for accelerated phase commonly used in clinical practice:

• WBC difficult to control with IFN and/or hydroxyurea in terms of dose required or shortening of intervals between course.

• >20% blasts or promyelocytes in blood or marrow

• >20% basophils/eosinophils

• Anemia and/or thrombocytopenia non-responsive to IFN

and/or hydroxyurea.

• Persistent thrombocytosis (>600 x 109/L)

• Additional chromosomal changes*

• Increasing splenomegaly unresponsive to treatment

• Unexplained fevers and bone pains

• Development of myelofibrosis or chloromas.

• This is the subject of ongoing debate, and some study protocols do not formally recognize this as a feature of accelerated phase disease.

Criteria for accelerated phase derived from multivariate analysis:

• Peripheral blasts plus peripheral blasts >15%

• Peripheral basophils >20%

• Thrombocytopenia (< 100 x109/L) unrelated to therapy

• Cytogenetic clonal evolution

Criteria for Blast Crisis:

• >30% blasts in peripheral blood and/or marrow

• Extramedullary infiltrate of blast cells presence of pi90BCR-ABL suggests de novo ALL and p210BCR-ABL might suggest pre-existing CML. It is possible, in both children and adults with CML, to develop blast transformation with mixed lineages, that is, both lymphoid and myeloid surface markers detectable on the same cells12 or some blasts with either lymphoid or myeloid characteristics.13 T-lymphoid blast transformation is rare but there are several cases showing both the BCR-ABL fusion gene and T-cel! receptor (TCR) gene rearrangements.14

Cytogenetic and molecular changes are well recognized in 50 to 80 percent of patients during transformation to accelerated or blast phase. Minor cytogenetic changes include monosomies of chromosomes 7 and 17, and loss of the Y chromosome;15 trisomies of chromosomes 17 and 21; and translocations of chromosome 3 with chromosome 21, t(3;21)(q26;q22).16 Major cytogenetic changes include trisomy 8, isochromosome i(17q), trisomy 19, and a double Ph chromosome.16-18

Alterations in p53 on the long arm of chromosome 17 by deletion, rearrangement, or mutation of sequences, occurring predominantly with myeloid blast crisis, have been distinguished in up to 30 percent of CML patients entering the blast phase.19,20 Even before their clinical manifestations it may be possible to detect these cytogenetic changes in the bone marrow, extramedullary masses, or splenectomy specimens.3,21,22

Clinical signs of blastic transformation may be due to the rapid exponential rise in blasts in the peripheral blood. The most significant areas compromised are the cerebral and respiratory circulations, resulting in multifocal bleeding, dyspnea, and hypoxemia.23,24 Tumors due to the deposition of blast cells, otherwise known as chloromas granulocytic sarcomas, may be visible before the detection of blasts in the peripheral blood.25 It is important to distinguish such tumors from undifferentiated carcinomas and diffuse large cell non-Hodgkin's lymphoma, which may require immnuohistochemical staining. Commonly the tumors are detected in lymph nodes, cutaneous tissue, or as lucent bone deposits on x-ray. Meningeal deposition may result in cord compression but has also been noted in patients who have achieved remission from blast transformation.26

Prognostic Scores

Prognostic models aim to categorize patients into different risk groups at diagnosis, which is particularly important in the analysis of trial data. These scores all require examination of the first blood film that is made on the newly diagnosed patient but regrettably the subsequent loss of this blood film is the most common reason for an incomplete score. Prognostic scores have traditionally been utilized in the context of analyzing large clinical trials and thus they should be applied to the individual patient with some caution. Small variations in the parameters described can make a significant difference to the final score; accurate determination of these parameters is therefore crucial. The details of the more commonly used systems are described in Table 5-4.

Tura et al.,27 and subsequently in 1982 Cervantes and Roz-man,28 devised scoring systems based on easily recordable parameters to categorize CML patients into three groups or stages. These were then applied to obtain the likely survival times

Table 5-4. The Prognostic Scores Commonly Used in CML

• Criteria required at presentation (prior to treatment):

2. Spleen size (cm below costal margin measured clinically with a tape)

3. Platelet count prior to any treatment

4. Blast percentage in peripheral blood (preferably 500 cells, but at least 200)

• Formula: Exp[0.0116(age - 43.4) + 0.0345(spleen - 7.51)

+ 0.188 (platelets/700) 2 - 0.563) + 0.0887 (% blasts - 2.1)] Good prognosis <0.8 Moderate prognosis 0.8 - 1.2 Poor prognosis >1.2

• An on-line calculator is available at: http://www.ncl.ac.uk/cml

Hasford score (1998) (Hasford, Pfirrmann et al.29)

• The 'Hasford' score is a more recent and refined, prognostic index

• Data analysed on 1573 patients

• Criteria required at presentation (prior to treatment):

2. Spleen size (cm below costal margin measured clinically with a tape)

3. Platelet count prior to any treatment

4. Blast percentage in peripheral blood (preferably 500 cells, but at least 200)

5. Eosinophil percentage in peripheral blood (same no. of cells counted)

6. Basophil percentage in peripheral blood (same no. of cells counted)

• Formula: (0.6666 x age [0 when < 50 years; otherwise 1]

+ 0.0420 x spleen size [cm from costal margin] + 1.0956 x platelet count [0 when platelets < 1500; otherwise 1] + 0.0584 x blasts [%] + 0.0413 x eosinophils [%]

+ 0.2039 x basophils [0 when basophils < 3%; otherwise 1] multiplied by 1000

• Interpretation:

• Low risk < 780 median survival 100 months

• Intermediate risk > 780 < 1480 median survival 69 months

• High risk >1480 median survival 45 months

• An on-line calculator is available at: http://www.pharmacoepi.de/cmlscore.html

Gratwohl's system (1998) (Gratwohl, Hermans et al.31)

• Criteria based on pre-transplant risk factors:

1. Histocompatibility (0 if HLA identical sibling; 1 if matched unrelated donor)

2. Stage of disease at time of transplant (0 if 1st chronic phase; 1 if accelerated phase; 2 if blast crisis or second or later chronic phase)

3. Age of donor and recipient (0 if <20 years; 1 if 20-40 years; 2 if >40 years)

4. Sex of donor and recipient (0 for all except 1 for male recipient/female donor)

5. Time from diagnosis to transplant (0 if <12 months; 1 if >12 months)

• Interpretation:

Score 5 year survival (%) Risk of transplant-related mortality (%)

0 72 20

1 70 23

2 62 31

3 48 46

4 40 51

5 18 71

6 22 73

for each stage, which differed significantly according to stage. The later models, such as the system devised by Sokal et al. in 1984,6,8 are based on multivariate analyses and the hazard ratio, derived using regression calculations. The Sokal model is the most extensively used in clinical research and is widely applied.

Hasford et al.29 devised a system based on a German trial population,30 which also includes the percentage of peripheral eosinophils and basophils. Patients can thus be stratified based on the presence or absence of these factors into three risk groups with significantly different outcomes.

Gratwohl et al.31 devised a simple system based on five main factors following analysis of 3142 patients allografted for CML between 1989 and 1997. These factors can be used for risk assessment to assist in the counseling of patients prior to allografting.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment