International Journal of Hematology

, 93:736

Utility of laboratory tests in B-CLL patients in different clinical stages

Authors

  • Małgorzata Rusak
    • Department of Hematological DiagnosticsMedical University of Bialystok
  • Joanna Osada
    • Department of Hematological DiagnosticsMedical University of Bialystok
    • Department of Hematological DiagnosticsMedical University of Bialystok
  • Joanna Chociej-Stypułkowska
    • Department of Hematological DiagnosticsMedical University of Bialystok
  • Milena Dąbrowska
    • Department of Hematological DiagnosticsMedical University of Bialystok
  • Janusz Kłoczko
    • Department of HaematologyUniversity Hospital in Bialystok
Original Article

DOI: 10.1007/s12185-011-0862-3

Cite this article as:
Rusak, M., Osada, J., Pawlus, J. et al. Int J Hematol (2011) 93: 736. doi:10.1007/s12185-011-0862-3

Abstract

The study objective was to analyse the utility of laboratory tests performed in 30 patients with B-cell chronic lymphocytic leukaemia (B-CLL) at different clinical stages. Laboratory tests included automated and microscopic assessment of peripheral blood and bone marrow counts as well as evaluation of leukaemic cells. Apart from the diagnostic and prognostic value of laboratory abnormalities such as clonal lymphocytosis with CD5+CD19+CD23+ phenotype, reduced erythrocyte parameters, thrombocytopenia or bone marrow infiltration by the neoplastic clone as well as low percentage of Gumprecht’s shadows, low apoptotic activity of peripheral blood lymphocytes, and increased percentage of CD38− and ZAP-70 ± cells markedly correlate with the stage of disease progression. These results seem to confirm the diagnostic and prognostic significance of these parameters determined in routine laboratory tests in B-CLL patients.

Keywords

B-CLLGumprecht’s shadowsApoptosisCD38ZAP-70

1 Introduction

Chronic lymphocytic leukaemia is the most common neoplastic disease of the haematopoietic system. Its average annual incidence rate is 2.4 cases per million population. It is characterized by proliferation and accumulation of lymphocytes arrested in the early phase of cell division. In over 90% of cases, B cells are involved. B-cell chronic lymphocytic leukaemia (B-CLL) accounts for 30–40% of all leukaemias diagnosed in Western Europe and in the United States of America, and is the most common type of leukaemia in patients over 65 years of life [1, 2]. It is twice as common in men than in women, with the annual morbidity rate ranging from 1 to 5.5/100,000 of the population. Until now, attempts have failed to identify a single disease-inducing agent, although B-CLL has been reported from farmers or workers at risk of rubber or asbestos exposure [3, 4].

Leukaemic cells in B-CLL probably originate from the population of CD5-positive B-1 lymphocytes and are characterized by prolonged survival due to disturbed apoptosis. This leads to the accumulation of monoclonal B cells in the G0/G1 phase in the peripheral blood and bone marrow and, as the disease progresses, also in lymph nodes, spleen and liver [5, 6]. It has been suggested that since neoplastic B cells, highly resistant to apoptosis in vivo, undergo this process spontaneously in vitro, not only abnormal intracellular mechanisms, but also some unidentified external factors may be among the causes that underlie their resistance to apoptosis [7]. The accumulation of lymphocytes is usually asymptomatic, the disease may go unnoticed for years and often it is diagnosed by chance. The diagnosis is based on morphological and genotypic features of the neoplastic cells and on the determination of prognostic factors [8]. Apart from history taking and physical examination, the diagnosis should involve certain laboratory investigations, such as determination of lymphocyte percentage and absolute count, shadows of disintegrated cells, immunophenotypic assessment of lymphocytes allowing identification of type of proliferation in the lymphatic system [7] and differentiation of B-CLL from non-neoplastic lymphocytosis. Lack of the proper differential diagnostics leads to over-diagnosis and unnecessary treatment in as many as 30% of patients with diagnosed B-CLL [9].

The characteristic feature of B-CLL is a heterogeneous course of the disease, despite similar clinical picture and morphology of the leukaemic cells. Clinical staging according to Rai or Binet classification and course dynamics (stable or progressive) are vital for making therapeutic decisions [8]. The varied clinical course of B-CLL generates the need to search for prognostic factors. In the evaluation of prognostic factors, in addition to imaging and cytogenetic diagnostics, also the degree of anaemia and thrombocytopenia in blood cell count as well as degree of bone marrow infiltration should be taken into consideration. The currently acknowledged prognostic factors also include the expression of CD38 on the surface of neoplastic B cells and cytoplasmic kinase ZAP-70 [10, 11].

2 Objective

The objective of this study was to analyse the diagnostic and prognostic value of laboratory tests based on (1) the results of routine haematological testing and immunophenotyping in B-CLL patients with various degrees of disease advancement, (2) the assessment of the correlation of the apoptotic activity of leukaemic lymphocytes with the presence of Gumprecht’s shadows and the expression of CD38 and ZAP-70 on the leukaemic cells at the time of diagnosis.

3 Materials and methods

3.1 Patients

The study involved 30 patients of Haematology Department, University Hospital in Białystok (20 men and 10 women aged 53–79 years, mean 63 years), with newly diagnosed B-CLL. In patients qualified for treatment, the investigations were performed prior to therapy implementation. The study was approved by the Bioethics Committee, Medical University of Białystok (R-I-002/78/2009).

Clinical staging was based on Rai or Binet classification. The patients were allocated in three groups of ten each:
  • group I/A: 2 women and 8 men (mean age 65.8 years);

  • group II/B: 5 women and 5 men (mean age 60.8 years);

  • group III, IV/C: 3 women and 7 men (mean age 62.5 years).

The control group consisted of 30 healthy subjects (20 men and 10 women) at the age of 32–81 years (mean age 62 years).

3.2 Materials and methods

The material for analysis included samples of peripheral blood and bone marrow collected into tubes containing K2-EDTA during routine diagnosis of B-CLL.

3.3 Complete blood count

Complete blood count was assessed using a haematological analyser Advia 2120 (Siemens).

The peripheral blood smear and bone marrow smear were assessed microscopically by means of May–Grünwald–Giemza staining [initial staining with May–Grünwald reagent (eosin, methylene blue, methanol, glycerol) for 5 min; after the pigment was removed, the preparation is stained with Giemza reagent (azure II, eosin, methanol, glycerol) for 25 min]. After rinsing and drying, the cells are evaluated under immerse magnification using light microscopy (1,000×). Leucocytes and the number of cell shadows (the so-called Gumprecht’s shadows) were assessed in peripheral blood preparations. In myelograms, the preparation density and the percentage of the respective cell lines were calculated.

3.3.1 Assessment of apoptosis intensity in isolated peripheral blood lymphocytes

The percentage of apoptotic cells was determined by cytometry, using double staining with Annexin-V-Fluos and propidine iodide (Becton-Dickinson, BD) immediately after collection (spontaneous apoptosis), and after 24-h incubation in serum-free RPMI 1640 (induced apoptosis). Results of the cytometric investigations were confirmed by the fluorescence microscopy method: isolated lymphocytes were stained with a mixture of acridine orange (10 μM) and ethidium bromide (10 μM) (Sigma), prepared in PBS. The mixture of stains (10 μl) was added to the suspension of cells (250 μl) and after 5 min. The stained cells were evaluated under a fluorescence microscope (200×)—100 consecutive cells [12]. The percentage of apoptotic cells was the mean calculated from the results obtained by the two research methods.

3.3.2 Flow cytometry

Immunophenotyping of lymphoid cells was performed using three-colour flow cytometry with a panel of antibodies: CD45, CD3, CD4, CD8, CD5, CD19, CD20, CD23 and CD45/CD14. Monoclonal antibodies conjugated with fluorochromes showing three emission ranges: fluorescein isothiocyanate (FITC), phycoerythrin (PE) and allophycocyanin (APC) were used to label the cells.

Negative controls were incubated with the fluorochrome-conjugated immunoglobulin, IgG1/G2a, having the identical isotype as the remaining antibodies. Cell antigens labelled with the above-mentioned antibodies were routinely analysed using a flow cytometer FACScalibur (BD), with a gate fixed on the CD45+/CD14− lymphoid cell population (10,000 cells/sample). It has been assumed that the cells show the antigen expression if the percentage of fluorescence-positive cells is higher than the isotopic control and is at least 20%.

The expression of CD38 on leukaemic B cells (CD19+CD5+) was analysed using anti-CD38PE monoclonal antibody (BD).

The presence of tyrosine kinase ZAP-70 was determined in leukaemic B cells with anti-ZAP-70PE antibody. The cut-off values, i.e. 30% for CD38 [1] and 20% for ZAP-70 [4], were defined as thresholds.

3.3.3 Statistical analysis

Results were analysed using the descriptive and analytical statistics. The descriptive statistics provides the characteristics of the studied variables by means of the arithmetic means, standard deviation and median. The analytical statistics aims to analyse differences and correlations between the study groups. For continuous variables, the Shapiro–Wilk test was used to check whether distributions are normal in the groups and for these the Student’s t test was applied. When the conditions of normal distribution were not fulfilled, the non-parametric Kolgomorov–Smirnov test was applied. The value of p < 0.05 calculated with SlideWrite was considered statistically significant. Correlations between the groups were evaluated by Pearson’s correlation coefficient (r) within the interval [0, 1] when correlation was positive and [0, −1] when it was negative. The value close to 1 indicates full correlation, whereas the value approaching 0 suggests that the correlation is weak. Data processing was performed using a package of MS Office 2003 (XP) Professional: MS Excel 2003, MS Word 2003, and statistical packages available at http://www.physics.csbsju.edu/stats/.

4 Results

4.1 The analysis of erythrocytic parameters, platelet count, leucocyte count and percentage of Gumprecht’s shadows and myelogram

The mean values of RBC, HGB and HCT were significantly reduced only in group III, IV/C (Table 1). Likewise, the mean platelet count in group I/A and group II/B was within reference limits (130–350 × 109/l), being 167.8 ± 36.82 × 109 and 161.75 ± 40.84 × 109/l, respectively. The mean platelet count was significantly decreased (p < 0.05) only in group III, IV/C (126 ± 157 × 109/l).
Table 1

Number of red blood cells (RBC), haemoglobin (HGB), the volume of haematocrit (HCT), platelet (PLT), white blood cell count (WBC), absolute and relative number of lymphocytes and neutrophils (Neut and Lym), the percentage of shadows Gumprecht (CG) and result myelogram (Lymp. system total lymphatic system total, Lymp. blasts lymphoblasts, Prolymp. prolymphocytes, Lymp. lymphocytes, Gran. system granulocytic system, Eryt. system erythrocyte system, Thro. system thrombocytic system)

Parameter

Controla

Classification Rai/Binet [mean ± SD (median)]

p

I/A

II/B

III, IV/C

Age

62 ± 1.2 (58)

65.8 ± 2.5 (63.5)

60.8 ± 4.8 (59)

62.5 ± 1.9 (61)

 

RBC (1012/l)

K:4–5.5, M:4.5–6 (4.89)

4.10 ± 0.64 (4.32)

4.20 ± 0.74 (3.72)

3.81 ± 0.98* (3.96)

*k:III, IV/C < 0.05

HGB (g/dl)

K:12–16, M:14–18 (15.1)

12.37 ± 2.14 (13.7)

12.52 ± 2.21 (11.9)

11.52 ± 2.46* (11.75)

*k:III, IV/C < 0.002

HCT (%)

K:37–41, M:40–54 (44.51)

37.52 ± 5.99 (39.05)

37.17 ± 6.37 (36.2)

34.61 ± 2.46* (35.7)

*k:III, IV/C < 0.009

PLT (×109/l)

130 × 350 (221.3)

167.8 ± 36.82 (173.5)

161.75 ± 40.84 (112.5)

123.6 ± 157* (83.0)

*k:III, IV/C < 0.05

WBC (×109/l)

4–101 (6.8)

118.98 ± 95.57* (89.0)

97.58 ± 43.81* (84.1)

121.23 ± 93.15* (94.9)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

Lym (×109/l)

0.9–4.5 (3.8)

107.28 ± 88.61* (74.14)

86.00 ± 40.62* (74.82)

107.72 ± 84.40* (80.16)

*k:I/A < 0.004

*k:II/B < 0.004

*k:III, IV/C < 0.004

Lym (%)

18–48 (38)

89.21 ± 5.31* (91.52)

86.68 ± 8.28* (90.43)

86.58 ± 8.86* (88.74)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

Neut (×109/l)

1.9–7.5 (6.5)

4.02 ± 2.28 (4.56)

7.13 ± 5.27 (5.83)

6.66 ± 3.51 (6.03)

ns

Neut (%)

40–72 (65)

4.03 ± 2.66* (3.0)

7.78 ± 4.26* (7.52)

8.03 ± 7.30* (6.2)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

C.G. (%)

0–1 (0.5)

47.3 ± 11.89*,** (43.5)

8.9 ± 5.32*,** (9.0)

2.2 ± 1.87** (2.0)

*k:I/A < 0.001

*k:II/B < 0.001

**I/A:III, IV/C < 0.001

**II/B:III, IV/C < 0.05

Lymp. system total (%)

3–12 (7.8)

67.86 ± 7.03* (67.27)

80.89 ± 5.18*,** (79.9)

84.20 ± 4.03*,*** (83.28)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:II/B < 0.003

***I/A:III, IV/C < 0.001

Lymp. blasts (%)

0

1.96 ± 1.3* (1.3)

7.54 ± 2.1*,** (8.3)

8.9 ± 1.4*,*** (8.8)

*k:I/A < 0.003

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:II/B < 0.002

***I/A:III, IV/C < 0.001

Pro-lymp. (%)

0

5.47 ± 1.5* (5.1)

10.2 ± 1.4*,** (9.9)

12.5 ± 1.6*,***,**** (12.0)

*k:I/A < 0.003

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:II/B < 0.001

***I/A:III, IV/C < 0.001

****II/B:III, IV/C < 0.008

Lymp. (%)

3–12 (7–8)

60.35 ± 7.1* (60.1)

63.1 ± 5.1*,** (63.4)

62.79 ± 4.3*,*** (62.05)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:II/B < 0.003

***I/A:III, IV/C < 0.001

Gran. system (%)

60–72 (65.4)

15.28 ± 8.75* (15.2)

7.50 ± 3.57* (6.18)

5.89 ± 4.43* (5.7)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

Eryt. system (%)

10–30 (22.8)

16.61 ± 7.01 (18.55)

11.12 ± 4.67* (10.9)

9.74 ± 4.26*,** (9.5)

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:III, IV/C < 0.05

Thro. system (%)

0–0.5 (0.3)

0.26 ± 0.15 (0.3)

0.14 ± 0.16 (0.11)

0.19 ± 0.21 (0.13)

ns

Values in parentheses are median values

I/A, II/B, III/C clinical staging was based on Rai or Binet classification

All the B-CLL patients, irrespective of the disease stage, had significantly increased leucocyte count (range from 25.81 to 327.2 × 109/l) as compared to the values observed in the control group (4–10 × 109/l, p < 0.001, Table 1).

Analogically, the relative and absolute lymphocyte counts in the peripheral blood of B-CLL patients were significantly higher in all the groups (I/A, II/B, III, IV/C: 107.23 ± 88.61 × 109, 86 ± 40.62 × 109, 107.72 ± 84.4 × 109/l and 89.21 ± 5.31, 86.68 ± 8.28, and 86.58 ± 8.86%, respectively) in comparison with the group of healthy subjects (absolute 0.9–4.5 × 109/l, relative 18–48%; Table 1).

The increase in lymphocytosis was accompanied by a decrease in the percentage of neutrophils in 100% of B-CLL patients as compared to the control group (the reference value 40–72%), irrespective of B-CLL stage. However, the absolute count of neutrophils was within the normal limits (1.9–7.5 × 109/l) in all the groups of patients (Table 1).

The percentage of cell shadows (the so-called Gumprecht’s shadows) calculated per 100 leucocytes was significantly higher (p < 0.005) in group I/A as compared to group II/B and group III, IV/C (47.3 ± 11.89 vs. 8.9 ± 5.32 vs. 2.2 ± 1.87%, respectively). Significant differences were also noted between group II/B and group III, IV/C (p < 0.05, Table 1).

Cell-rich and pronouncedly cell-rich myelograms predominated in all the study groups (Table 1). In all the patients, the lymphatic system significantly exceeded the reference value (3–12%, p < 0.001). There were significant differences in the percentage of lymphoblasts (1.96 ± 1.3 vs. 7.54 ± 2.1 vs. 8.9 ± 1.4) and prolymphocytes (5.47 ± 1.5 vs. 10.2 ± 1.4 vs. 12.5 ± 1.6) between the study groups, indicating a directly proportional correlation with B-CLL stage (Table 1).

The granulocytic system was attenuated in all the B-CLL patients, with the lowest values noted in group III, IV/C (5.89 ± 4.43%, Table 1).

The erythrocyte system was within the lower reference limit (10–30%) in group I/A and group II/B (16.61 ± 7.01 and 11.12 ± 4.67%, respectively). In group III, IV/C, the percentage of the red blood cells was significantly lowered as compared to group I/A (9.74%; p < 0.05).

The thrombocytopoietic system was within the reference limits in all the study groups (Table 1).

4.2 Flow cytometry and assessment of lymphocyte apoptosis

In all the study groups, the expressions of CD19 and CD5/CD19 antigens were significantly higher as compared to the control value (p < 0.001). In group I/A, the expressions of CD19 and CD5/CD19 were significantly lower in comparison with groups II/B and III, IV/C (80.31 ± 6.32 and 74.17 ± 5.54% vs. 87.61 ± 7.27 and 82.78 ± 9.44% vs. 88.33 ± 4.07 and 82.9 ± 4.26%, respectively, Table 2).
Table 2

Evaluation of immunophenotype cytometry and the percentage of cells apoptotic in peripheral blood

Parameter

Controla

Classification Rai/Binet [mean ± SD (median)]

p

I/A

II/B

III, IV/C

CD45 (%)

100 (98.7)

98.52 ± 0.83 (98.39)

98.41 ± 0.82 (98.34)

98.83 ± 0.46 (98.79)

ns

CD19 (%)

7–23 (17.8)

80.31 ± 6.32* (80.71)

87.61 ± 7.27*,** (89.49)

88.33 ± 4.07*,*** (87.87)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0

001

**I/A:II/B < 0.02

***I/A:III, IV/C < 0.008

CD5/19 (%)

1–2 (1.2)

74.17 ± 5.54* (74.50)

82.78 ± 9.44*,** (84.99)

82.90 ± 4.26*,*** (81.05)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:II/B < 0.03

***I/A:III.IV/C < 0.001

CD3/CD4 (%)

28.5–60.5 (38.1)

3.42 ± 3.23* (2.12)

2.17 ± 0.91* (2.45)

3.14 ± 1.12* (3.57)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III.IV/C < 0.001

CD3/CD8 (%)

11.1–38.3 (27.6)

2.58 ± 1.93* (2.04)

2.01 ± 0.67* (2.1)

3.48 ± 0.54* (3.41)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

CD23 (%)

2–4 (2.7)

69.40 ± 9.36* (69.75)

73.76 ± 11.3* (77.56)

70.47 ± 14.4* (73.20)

*k:I/A < 0.0001

*k:II/B < 0.0001

*k:III, IV/C < 0.0001

CD38 (%)

5–7 (6.1)

7.92 ± 3.42 (7.84)

43.5 ± 14.4*,** (40.46)

69.6 ± 16.4*,***,**** (66.01)

*k:II/B < 0.001

*k:III.IV/C < 0.001

**I/A:II/B < 0.001

***I/A:III, IV/C < 0.001

****II/B:III, IV/C < 0.009

ZAP-70 (%)

1–2 (1.4)

9.88 ± 4.00* (8.45)

55.13 ± 15.14*,** (52.42)

63.8 ± 17.24*,*** (60.27)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:II/B < 0.001

***I/A:III, IV/C < 0.001

Apoptosis spontaneous (%)

5 ± 2 (3)

11.3 ± 2.5* (11.0)

9.8 ± 4.5* (9.5)

9.2 ± 2.2* (9.5)

*k:I/A < 0.001

*k:II/B < 0.005

*k:III, IV/C < 0.001

Apoptosis induced (%)

11 ± 4 (9)

43.9 ± 4.0*,** (43.5)

40.5 ± 5.6*,** (40.0)

32.3 ± 4.88*,** (30.5)

*k:I/A < 0.001

*k:II/B < 0.001

*k:III, IV/C < 0.001

**I/A:III.IV/C < 0.001

**II/B:III, IV/C < 0.001

Values in parentheses are median values

I/A, II/B, III/C clinical staging was based on Rai or Binet classification

In all the study groups, irrespective of B-CLL stage, CD23 antigen expression was abnormally high (I/A: 69.40 ± 9.36%, II/B: 73.76 ± 11.3%, III, IV/C: 70.47 ± 14.4%), whereas there were no antigens that are present on normal T cells (CD3, CD4, CD8; Table 2).

The expressions of CD38 and ZAP-70 were significantly increased as compared to the control values (p < 0.001). Significant differences were noted in the expression of both antigens, depending on the advancement stage of the neoplastic process. In group II/B, the expressions of CD38 and ZAP-70 were significantly higher than in I/A (43.5 ± 14.4 vs. 7.92 ± 3.42% and 55.13 ± 15.14 vs. 9.88 ± 4%, respectively). In group III, IV/C, the expression values were even higher: CD38 69.6 ± 16.4% and ZAP-70 63.8 ± 17.24% (Table 2).

The enhancement of spontaneous and induced lymphocyte apoptosis in healthy subjects was 5 ± 2 and 11 ± 4%, respectively (Table 2). In all the study groups, the percentage of lymphocytes showing spontaneous and induced apoptosis was significantly increased, and a distinctly reversely proportional correlation occurred between apoptosis intensity and the disease stage (I/A: 11.3 ± 2.5%/43.9 ± 4%; II/B: 9.8 ± 4.5%/40.5 ± 5.6%; III, IV/C: 9.2 ± 2.2%/32.3 ± 4.8%; Table 2).

4.3 Correlations

A high positive correlation was observed between the percentage of spontaneous and stimulated apoptotic cells and the percentage of Gumprecht’s shadows between the sum of the results obtained in the test groups (0.990 vs. 0.841, respectively; Fig. 1a, b).
https://static-content.springer.com/image/art%3A10.1007%2Fs12185-011-0862-3/MediaObjects/12185_2011_862_Fig1_HTML.gif
Fig. 1

Correlations between the percentage of apoptotic cells, Gumprecht shadow, expression of CD38+ and ZAP-70 (I/A, II/B, III/C- clinical staging was based on Rai or Binet classification)

In the same groups, a statistically significant negative correlation was found of the percentage of apoptotic cells with CD38+ (−0.988 vs. −0.948; Fig. 1c, d) and with ZAP-70+ expression (−0.821 vs. −0.821; Fig. 1e, f).

5 Discussion

The analysis of routine haematological investigations has confirmed that elevated leukocytosis with relative and absolute lymphocytosis is the most sensitive diagnostic parameter in B-CLL, irrespective of the stage of the neoplastic disease. These alterations are observed already in the early stage, when the percentage involvement of the lymphatic system in the bone marrow exceeds the reference value. The analysis has also indicated the proportional correlation of B-CLL stage with the counts of lymphoblasts and prolymphoblasts in the myelogram. Laboratory indices of anaemia and thrombocytopenia appeared only in the most advanced stages. In the study group, anaemia seems to be the result of dislodgment of the erythroblastic system by a neoplastic lymphocyte clone, whereas the normal count of megakaryocytes indicates the peripheral and immune character of thrombocytopenia. Despite suppression of the granulocyte system in the bone marrow, absolute neutropenia was not observed even in the most advanced stage of B-CLL. An interesting observation refers to the cell shadows (the so-called Gumprecht’s shadows) which until recently were referred to as a laboratory artefact, produced during preparation of peripheral blood smear in patients with B-CLL. Thus, traditionally, the presence of Gumprecht’s shadows was considered to be a laboratory feature, pathognomic for chronic lymphocytic leukaemia. However, as revealed in our study, the quantitative evaluation of this phenomenon may have a significant diagnostic and prognostic value, as the percentage of Gumprecht’s shadows strongly correlates with B-CLL stage and is the highest in patients with better prognosis. Moreover, we observed a markedly positive correlation between the Gumprecht’s shadows and the percentage of apoptotic lymphocytes. It has been known that lymphocytes in B-CLL show an increased expression of the antiapoptotic oncoprotein bcl-2, which is associated with resistance of leukaemic lymphocytes to apoptosis and in consequence, with prolonged survival and hindered removal from the body [9, 13]. In our study, most apoptotic lymphocytes, both spontaneous and induced, were observed in patients with the best clinical condition and with the largest number of Gumprecht’s shadows. It can be thus assumed that Gumprecht’s shadows are remnants of destroyed leukaemic B cells, and that small numbers of Gumprecht’s shadows or apoptotic cells may be an unfavourable prognostic factor in the course of B-CLL. The correlation of higher percentage of Gumprecht’s shadows with better prognosis and longer survival of patients with chronic lymphocytic leukaemia has also been confirmed by Nowakowski [14].

The leukaemic lymphocytes present in B-CLL probably originate from the CD5-positive B-1 lymphocytes that appear in early ontogenesis and after week 17 of gestation from the main group of B cells in the spleen and lymph nodes. In the umbilical blood, they account for as many as 50–95% of cells. In adults, the antigen CD5+ is characteristic of T lymphocytes and occurs in 5–30% of peripheral blood B cells. The CD5+ cells account for approximately 10% of cells in the spleen and 30% of the mantle zone of lymph nodes in the tonsils. As shown by literature data, the primary neoplastic transformation may result from continuous activation of CD5+ B cells by autoantigens [2]. In the immunophenotypic investigations performed in our centre, the leukaemic cells were characterized by DC5+ and CD23 expressions on the CD19+ cells and lack of CD3, CD4 and CD8. Such disorders are not observed on the reactive lymphocytes. High expression of CD5 antigen allows exclusion of hairy cell leukaemia and in 50% prolymphocytic leukaemia, in which this antigen is negative. The presence of CD23 on the surface of the studied cells confirmed the diagnosis of B-CLL and excluded mantle zone lymphoma, in which this antigen is not found. The number of CD5+CD19+CD23+ cells was observed to increase significantly with B-CLL advancement [15].

Similar observation referred to the cells showing CD38 expression associated with Ig (V)H mutation [2, 6]. In patients with better prognosis, the percentage of CD38+ cells was comparable to the group of healthy subjects, whereas the number of positive cells increased significantly with the disease stage. It is claimed that CD38+ patients exhibit higher telomerase activity and increased percentage of cells positive for Ki-67 that labels the proliferating cells. It has been also demonstrated that CD38+ expression manifests activation of the signalling pathway that leads to the proliferation of leukaemic B cells, their resistance to proapoptotic stimuli and thus to the increased aggressiveness of B-CLL [4]. This has also been confirmed by our investigations, in which the percentage of CD38+ cells was found to correlate strongly negatively with lymphocyte sensitivity to apoptosis. These findings are also consistent with the observation that the percentage of CD38+ cells lower by 20% indicates better prognosis [9, 14] and treatment prognostication [10]. According to Ibrahim et al., determination of CD38 expression allows identification of patients with potentially progressive nature of leukaemia irrespective of the clinical condition. It should be remembered that CD38 expression tends to fluctuate in the course of the disease [9, 16].

The increased expression of the molecule CD38 in the leukaemic B cells is accompanied by the presence of the ZAP-70 antigen, being related to the increased signalling activity that occurs through the BCR receptor [10]. The convenient diagnostic feature of ZAP-70 is the fact that, unlike CD38, its expression at the disease onset remains stable throughout its course. It is believed that high level of ZAP-70 indicates rapid progression of lymphocytic leukaemia and reduction in the mean survival time, irrespective of the disease advancement at the time of treatment implementation [17, 18]. In our group of low-stage patients, fewer than 10% of cells were ZAP-70 positive. In groups with II/B and III, IV/C stage, the mean percentage of ZAP-70 positive cells exceeded 50%. According to literature data, the presence of over 30% of ZAP-70 positive cells indicates ZAP-positive B-CLL (the so-called non-mutated), characterized by worse prognosis. The percentage of such cells is estimated at around 40% of all B-CLL cases studied [1, 17]. Our findings also demonstrated a strongly negative correlation of both CD38 and ZAP-70 expression on cells in B-CLL with lymphocyte vulnerability to apoptosis.

Apart from high clonal lymphocytosis, anaemia and thrombocytopenia, also a small number of Gumprecht’s shadows, low apoptotic activity and high percentage of CD38 and ZAP-70 positive cells may serve as prognostic factors that are complementary to one another and that may help identify patients with an unfavourable course of B-CLL and response to treatment already at an early stage of neoplasm advancement.

6 Conclusions

The current laboratory study conducted in patients with chronic B-cell lymphocytic leukaemia in different disease stages seems to emphasize the role of leucocyte immunophenotyping as an indispensable tool in the diagnosis of B-CLL. The expression of the so-called primary B-CLL antigens allows differentiation of neoplastic B-lymphocytes from polyclonal proliferation and other lymphatic proliferations. The quantitative evaluation of cells with positive CD38 and ZAP-70 expressions is a significant prognostic factor. Also the quantitative evaluations of Gumprecht’s shadows in the peripheral blood smear and vulnerability to spontaneous or induced apoptosis, showing higher values in patients with better prognosis, have both diagnostic and prognostic utility.

Copyright information

© The Japanese Society of Hematology 2011