Immunophenotypic analysis of T-cell acute lymphoblastic leukaemia in Madras, India

Leukemin R~s~nrch Vol. 21, No. 2, pp. 119-124, 1997. Copyright Q 1997 Elscvier Science Ltd. All rights rcscwed Prmted m Great Bnfam 014552126/97 $17.W + 0.00

Pergamon PII: SO1452126(96)00092-6

IMMUNOPHENOTYPIC ANALYSIS OF T-CELL LYMPHOBLASTIC LEUKAEMIA IN MADRAS, Kamalalayam

Raghavan

ACUTE INDIA

Rajalekshmy *, Arcot Radhakrishnan Abitha*, Radhakrishnan Tenali Gnanasagart and Viswanathan ShantaJ

Pramila*,

*Haematology/Immunology Division, Cancer Institute (W.I.A), Adyar, Madras 20, Pin Code 600 020, South India; tMedica1 Oncology Division, Cancer Institute (W.I.A), Adyar, Madras 20, Pin Code 600 020, South India; and ICancer Institute (W.I.A), Adyar, Madras 20, Pin Code 600 020, South India (Received

5 August

1996. Accepted 24 August 1996)

Abstract-lmmunophenotypic analysis of 285 newly diagnosed previously untreated, unselected, acute lymphoblastic leukaemia cases carried out at the Cancer Institute (W.1.A) in Madras reveals that 126 (44.2%) cases showed T-immunophenotype. The study was conducted using flow-cytometric immunofluorescent or immunoperoxidase methods using an extensive panel of monoclonal antibodies comprising CDI, 2,3,4,5,7,8,57, 19,20 K, I, IgG, M, D, Clg, CDIO, HLA-DR, CD13, 14, 33, 34 and CD61. The study group comprised 73 (57.9%) paediatric cases (< 15 years) and 53 (42.1%) adult cases (>I5 years). Based on their reactivity with various anti-T-cell monoclonal antibodies, all T-ALL cases were assigned to one of the intrathymic differentiation compartments, 56.2% of paediatric T-ALLs arise from intrathymic compartment II, 34.2% from compartment III and 9.6% from compartment I. Among adults, 45.3% arise from compartment I. 33.9% from compartment III and 20.8% from compartment II. The most frequently observed CD antigens in our study group are CD7,5,2 and 3. A correlative study of socioeconomic status reveals that 67.5% of T-ALL cases occurred among lower socioeconomic strata. 0 1997 Elsevier Science Ltd. All rights reserved. Key words:

T-ALL,

immunophenotypic

analysis,

Cancer

Institute

(WIA),

Madras.

21.4% adult cases were of C-ALL phenotype. This is at variance with Western reports where the immunophenotype is predominantly C-ALL. This finding raised the possibility that phenotypes of T-cell malignancies also may differ from their Western counterparts. T-cell malignancies have been extensively studied in Western countries with regard to their distribution according to the classical intrathymic differentiation compartments. T-ALL cases have been reported to number of cases originating from include a significant compartment I thymocytes. In some reported series, this has been a predominant phenotype [2]. Lymphoblastic lymphoma, on the other hand, originates predominantly from compartments II and III and rarely from compartment I thymocytes [3]. T-ALL also has been reported to be extremely heterogeneous with regard to the expression of various T-cell antigens [3-51. In addition, monoclonal antibodies reactive with the T-cell lineage are known to have different frequencies of reactivity with T-ALL cases. CD5 and CD7 have been claimed by several authors to be reactive with a high proportion of T-ALL

Introduction Immunophenotypic analytical study of Acute Lymphoblastic Leukaemia cases carried out during the year 1993 on 125 cases revealed that there is a predominance of T-ALL and a paucity of C-ALL in our population [l]. Hence, a detailed analytical study of T-ALL accrued till date (December 1995) was carried out to determine whether qualitative and/or quantitative differences from series reported from Western countries exist in our series. Immunophenotypic analysis of 285 cases shows that 46% of paediatric ALLs and 42.1% adult ALLs showed T-immunophenotype. 24.5% of paediatric cases and

Abbreviations: CD, cluster designation; ALL, acute lymphoblastic leukaemia; FAR, French-American-British; ACP, acid phosphatase; PAS, periodic acid Schiff; NSE, Non-specific esterase; C-ALL, common acute lymphoblastic leukaemia; LBL, lymphoblastic lymphoma; Gm%, haemoglobin in gram per 100 ml of blood; BD, Becton Dickinson. Correspondence to: V. Shanta, Director and Scientific Director, Cancer Institute, Adyar, Madras-20, South India, Pin: 600 020.

1691. 119

120

K. R. Rajalekshmy et

al.

Table 1. Distribution of ALL immunophenotypesat the Cancer Institue (W.1.A) (age- and sex-specific, 285 cases) Paediatric (< 1.5years) Male

Female

Adult (>15 years) Total

Male

Female

Total

Total

Immunophenotype

NO

%

NO

%

NO

%

NO

%

NO

%

NO

%

NO

%

T-ALL C-ALL Progenitor B PrecursorB Pre B B-cell Hybrid Unclassified

50

54.9 18.7

23 22

33.8 32.3

73 39

45.9 24.5

45 20

45.5 20.1

8 7

29.6 25.9

53 27

42.1 21.4

126

0.6 10.1

-

11.1

-

3

9 11

7.1 8.7

44.2 23.2 0.3 8.9

2

1.6

Total

17

-

-

1

1.5

1

8 2 4 7 3

8.8 2.2 4.4 7.7 3.3

8 8 2

16

3

11.8 11.8 2.9 1.5 4.4

6 8 6

6.3 3.8 5.0 3.8

91

57.2

68

42.8

159

55.8

1

10

Materials and Methods Immunophenotyping was carried out on 285 newly diagnosed, untreated, unselected ALL cases using an extensive panel of monoclonal antibodies. All cases met conventional morphological and cytochemical criteria for the diagnosis and were further classified by FAB system [lo]. Cases phenotyped as T-cell were subjected to detailed analysis. Cases included a total of 126 TALL. Out of this, 73 were children and 53 were adults. There were 95 males and 31 females (Table 1). Cell preparation Leukaemic cells were obtained mostly from bone marrow. In some cases, peripheral blood with more than 90% of circulating blast cells was used for the study. The mononuclear cell layer was separated on Ficollhypaque gradient [ 111. First, 125 cases were studied using a flow-cytometric immunoperoxidase method in haematology autoanalyser Hl system (Bayer, County Dublin, Ireland) [l]. The remaining 160 cases were studied using a flow-cytometric immunofluorescent method on FACScan (Becton Dickinson, San Jose, California), directly or indirectly, using the single- or dual-colour technique [12]. Monoclonal antibodies The monoclonal antibodies used, and their sources, are as follows: CD1 (Leu6), CD2 (LeuSB), CD3 (Leu4) CD4 (Leu3), CD5 (Leul), CD7 (Leu9) CD8 (Leu2), CD57 (Leu7) CD19 (Leul2), CD20 (Leu16) K, Iti, cytoplasmic u, CD10 (CALLA), HLA-DR, CD13 (LeuM7) CD14 (LeuM3), CD33 (LeuM9), CD34, CD61 The specifications of the monoclonal antibodies used are presented in Table 2.

-

66

6

6.1

7

7.1

4

1

1.0

1

14 6

14.1 6.1

3

1

11.1 3.7

17 7

13.5 5.6

1 25 21 8 25 13

99

78.6

27

21.4

126

44.2

285

14.9 3.7

7.4 2.8

8.8 4.4 100

Criteria for scoring a case as positive When the percentage of blast cells in the test sample exceeded 90%, a positive reaction was recorded if 20% or more of the cells reacted with a given monoclonal antibody. In cases with a lower percentage of blasts, the proportion of positive cells in the population required for the sample to be called positive was based on the percentage of blast cells to ensure that the reactivity of admixed normal cells did not lead to any erroneous conclusions regarding the phenotype of the leukaemic cells. This was especially important with the Tphenotype. Definition of immunophenotype The phenotype of a given case was characterized according to its reactivity with various monoclonal antibodies as shown in Table 3. Criteria for classification according to intrathymic compartments The system used in this work to assign cases to a given intrathymic compartment according to the pattern of reactivity with monoclonal antibody is shown in Table 4. Results Of the 285 cases immunophenotyped, 126 cases were classified as T-ALL. Ninety-five were males (75.4%) and 31 were females (24.6%). Seventy-three (57.9%) were children, and 53 (42.1%) were adults (Table 1). As shown in Table 5, 56.2% of paediatric T-ALLs arise from intrathymic compartment II, 34.2% from compartment III and 9.6% from compartment I. Among adults, 45.3% arise from compartment I, 33.9% from compartment III and 20.8% from compartment II. The most frequently positive T-cell antigens in our TALL series were CD7 (90%) CD5 (80%), CD2 (73%)

121

Phenotypic analysis of T-ALL in Madras, India Table 2. Specificity of monoclonal antibodies used T-cell CD1 (Leu6) CD2 (LeuSB) CD3 (Leu4) CD4 (Leu3) CD5 (Leul) CD7 (Leu9) CD8 (Leu2) CD57 (Leu7)

Cortical thymocytes Pan T-lymphocytes (sheep erythrocyte receptor) Pan T-lymphocytes T-helper cells Pan T-lymphocytes, Pan thymocytes Pan T-lymphocytes T-suppressodcytotoxic Natural killer cells

B-Cells CD19 (Leul2) CD20 (Leul6) Kappa Lambda W, M, D Cytoplasmic p

B lymphocytes B lymphocytes Kappa light chain Lambda light chain Heavy chain

Others CD10 (CALLA) HLA-DR CD34 CD61

Common ALL antigen B-cells, activated T-cells Progenitor cells Platelets

Myeloid markers CD13 (LeuM7) CD14 (LeuM3) CD33 (LeuM9)

Monocytes and myeloid cells Monocytes and myeloid cells Monocytes and myeloid cells

Table 3. Criteria followed for classification of ALLs into immunological subtype Phenotype T Precursor B C-ALL Pre B B Unclassified Hybrid

Markers that must be negative

Markers that must be positive At least one T-cell marker (CD2, 5 or 7) CD19, HLA-DR CDlO, CD19, HLA-DR CIg K/I, CD19, CD20, HLA-DR Lymphoid and myeloid markers or T-cell and B-cell markers

B-cell markers T markers, Kappa Lambda, Clg T, ~11, Clg T, K//? T All except HLA-DR

and CD3 (65%). CD4 and CD8 were positive in 39% of T-ALL. Correlation with FAB classification and cytochemical study shows that 84.2% T-ALLs are of L2 morphology and 15.8% are of Ll morphology. 73.3% of T-ALLs showed ACP positivity, 33.3% PAS positivity and 13.3% NSE positivity. The socioeconomic statuses of these patients were studied, taking into consideration the total annual income of the family, level of education, housing and sanitary conditions. Studies reveal that, among 285 acute lymphoblastic leukaemia cases, 183 (64.2%) were from lower-, 28% from middle- and 7.8% from upper socioeconomic strata. A socioeconomic status study conducted by the Madras Metropolitan Tumour Registry on the general population reveals that 61%, 33% and 6% of people

Variable markers Other T-cell markers HLA-DR CD20 CD10 HLA-DR

belong to lower-, middle- and upper socioeconomic groups, respectively. Hence the data on socioeconomic profiles of ALL cases are comparable with those of the general population. Eighty-five out of 126 T-ALL cases (67.5%) originated from among the lower socioeconomic groups. Among lower socioeconomic groups, the predominance of T-ALL was found to be statistically significant (P < 0.001). In our study series, paediatric T-ALLs arise at an older age group, with 38.6% in the 3-lo-year-old group and 59.6% in the >lO-year-old group. There is a distinct male predominance; 75.4% of the cases are male patients, and there is a M:F ratio of 3:l. 22.8% of TALLs showed a total WBC count of 10,000~9,000/ mm3, whereas 42.1% showed a total WBC count of >50,000/mm3; 45.6% of cases showed a haemoglobin level below 7 Gm%, 31.6% showed 7-lOGm% and

122

K. R. Rajalekshmy et al.

Table 4. Criteria adopted for classification of T-Al& Compartment Positive I II III

into intrathymic compartments

Variable

Negative

CD2, CDS and/or CD7 CD2, CD5, CD7 CDl, CD3 and or both CD4 and CD8 CD2, CD5, CD7 CD4 or CD8 CD2, CD3, CD5, CD7

CDl, CD3, CD4, CD8 CD1

Table 5. Distribution of T-ALLs according to intrathymic compartments(126 cases) Paediatric (< 15 years) Intrathymic compartment

Adult (>15 years)

Total

NO

%

NO

%

NO

%

9.6 56.2 34.2

24

III

7 41 25

11 18

45.3 20.8 33.9

31 52 43

24.6 41.3 34.1

Total

73

57.9

53

42.1

126

I II

only 22.8% showed a haemoglobin level above 10 Gm%. The platelet count was as low as 25,000-50,000/mm3 in 61.4% of cases. Massive lymphadenopathy was observed in 68.4% cases. 77.2% showed massive hepatosplenomegaly, 28.1% showed mediastinal widening and 10.5% showed tonsillar enlargement. 19.3% of cases presented with lymphoma syndrome. Testicular enlargement and CNS involvement, at presentation, were seen in 3.5% and 1.8% of the cases, respectively.

Discussion Immunophenotypes of ALL have been identified in different geographic and ethnic groups. A comparison of relative frequencies and incidence of major immunosubtypes indicates that the relative predominance in TALL and the selective shortage of C-ALL seen in our geographical setting are quite different from those reported from Western countries [l, 14-171. Patterns in our series show more or less the same trends as those reported from other Eastern and African countries [1826]. Among our T-ALL cases, the majority of paediatric T-ALLs arise from compartment II (56.2%) and compartment III (34.2%). Only 9.6% arise from compartment I. This contradicts the early findings of Reinherz et al. [2], who reported that 80% of T-ALL cases that they tested were phenotypically consistent with compartment I thymocytes, while the remaining 20% were characterized as compartment II thymocytes. Greaves et al. [27] also reported that 45 out of 46 cases of ALL cases originated from compartment I, with the 46th case originating from compartment II. Such

100

cases may as well arise from bone marrow. Crist et al. [28] reported 34%, 43% and 23% of their paediatric ALL cases belonged to thymic compartments I, II and III, respectively. Roper et al. [3] reported that 33%, 37% and 30% of paediatric ALL cases originated from compartments I, II and III, respectively. From Egypt, Kamel et uE. [29] reported that 58% of paediatric TALLs expressed a phenotype of compartment II, which is in agreement with our findings. Cases with phenotypic expression of II and III probably arise in the thymus. In our series, we find that 45.3% of adult T-ALLs arise from thymic compartment I, 33.9% from compartment III and 20.8% from compartment II. The trend in our adult T-ALL series is the same as in Kamel et uZ.‘s series [29] where adult T-ALLs show a distribution of 33.3%, 40% and 26.7% in intrathymic compartments I, II and III, respectively. T-cell lymphoblastic lymphoma cases predominantly express the markers of thymocyte compartments II and III [3,30,31], particularly when patients with bone marrow involvement are excluded (confirming the thymic origin). One difficulty in interpreting the published findings is that the designation of an individual case as leukaemia or lymphoma at times can be arbitrary. Many paediatric oncologists, for example, diagnose lymphoma if the proportion of blast cells in bone marrow is less than 25%, and diagnose leukaemia if the percentage is greater than 25%. Since such a distinction is subject to a bone marrow sampling error [32], and, moreover, relapses in bone marrow occur frequently, it is unlikely that such a distinction is biologically meaningful. None the less, the findings of a high proportion of neoplasms with thymic compartments II and III characteristics in patients with extensive bone

Phenotypic analysis of T-ALL in Madras, India

marrow infiltration by blasts are consistent with the possibility that such cases may have presented later in their clinical course. Such patients might have been classified as LBL at an earlier point in their illness. If this explains our findings, it could also account for the relatively high proportion of ALLs with T-cell characteristics in our series; some of them could be LBL in the leukaemic phase (i.e. arising from the thymus, but spreading to the bone marrow). In addition, in developing countries, lymphoblastic neoplasms that involve the more mature thymocyte compartments are more likely to present as leukaemias, perhaps because of their true biological difference. Such a biological difference could arise because of differences in socioeconomic status, and, if true, might also account for differences in the relative proportion of cases in late thymic compartments among the U.S. series. It should be noted that the series reported by Roper et al. was reported from Alabama, and that of Reinherz was from Boston. The fact that, in our series, 64.2% of T-ALL had their origin in lower socioeconomic groups might be very significant in this context. Whether or not there is a true paucity of early thymocyte neoplasms in patients of low socioeconomic status, blacks or patients in less developed countries, or an increase in late thymocyte neoplasms is an issue warranting further study. In our study, CD7 and CDS were found to be the most frequently reacting monoclonal antibodies with T-ALL case. This finding is in agreement with Western reports, in which the most consistently acting markers of T-ALL cases were found to be CD7 [6,7,33,34], and the report from Egypt [29] in which the most frequently reacting monoclonal antibodies were CD5 and then CD7. The male predominance in T-ALL and the prevalence in the occurrence of T-ALL among the older age group seen in our series are reported in other parts of the world [19, 291.

Conclusion We have demonstrated that 44.2% of newly diagnosed ALL cases (126/285) show a T-immunophenocases showed a distinct male type. T-ALL predominance, being more prevalent in older children. Paediatric T-ALL cases predominantly originated from more mature intrathymic compartments II and III. The most frequently reacting CD antigens were CD7 and CD5, followed by CD2 and CD3. 67.5% of T-ALL cases had originated from lower socioeconomic groups, suggesting a link between socioeconomic status and a prevalence of T-ALLs in a particular geographic setting. Further studies are warranted to confirm this possibility.

123

References 1. Rajalekshmy K. R., Abitha A. R., Premila R., Sagar T. G., Maitreyan V. & Shanta V. (1994) Immunophenotyping of Acute Lymphoblastic Leukemia in Madras, India. Leukemia Res. 18, 183. 2. Reinherz E. L., Kung P. C., Goldsteen G., Levey R. H. & Schlossman S. F. (1980) Discrete stages of human intrathymic differentiation, Analysis of normal thymocyte and leukemic lymphoblasts of T-cell leneage. Proc. natn. Acad. Sci. U.S.A. 77, 1588. 3. Roper M., Christ W. M., Metzgar R., Ragab A. H., Smith S., Starling K., Pullin J., Leventhal B., Bartolucci A. A. & Cooper M. D. (1983) Monoclonal antibody; characterisation of surface antigens in childhood T-cell lymphoid malignancies. Blood 61, 830. 4. Kozenir B., Gebhard D., Denny T., Mckenize S., Clarkson B. D., Miller D. A. & Evan R. L. (1982) Analysis of T cell differentiation antigens in acute lymphatic leukemia using monoclonal antibodies. Blood 60, 7.52. 5. Herrmen F., Komischke B., Sieber G. & Rehl H. (1983) Heterogenecity of T-cell neoplasia as defined by monoclonal antibodies. Klin. Wochensch. 61, 807. 6. Haynes B. F., Mann D. L., Helmer M., Schroer J., Schelhamer J., Eisenbarth G., Stromenger J., Thomas C., Mostwake H. & Fauci A. (1980) Characterization of a monoclonal antibody that defines an immuno regularity T cell subset for immunoglobulin synthesis in humans. Proc. natn. Acad. Sci. USA. 77, 2914. 7. Link M., Warnke R., Finlay J., Aamylon M., Miller R., Killey J. & Levy R. (1983) A single monoclonal antibody identified in T cell lineage of childhood lymphoid malignancies. Blood 62, 722. 8. Foon K. A., Schroff R. W. & Gale R. P. (1982) Cell surface markers on leukemia and lymphoma cells, recent advance. Blood 60, 1. 9. Schroff R. W., Foon K. A., Billing R. J. & Fahey J. H. L. (1982) Immunologic classification of lymphocytic leuke-

mias based on monoclonal antibody defined cell surface antigens. Blood 59, 207. 10. Bennet J. M., Catovsky D., Daniel M. J., Flandrin G., Galton D., Gratnick H. R. & Sultan C. (1981) The morphological classification of acute lymphoblastic leukemia concordance among observers and clonical correlations. Br. J. Haematol. 47, 553. 11. Boyum A. (1968) Separation of leukocytes from blood and bone marrow. Stand. J. Lab. Invest. 21, 97, 77. 12. Monoclonal Antibodies Source Book (1992) Becton Dickenson Immunocytometry system. 23-1429-06. 13. Becton Dickenson package insert cycle test plus DNA reagent kit for analysis of nuclear DNA from solid tissue or cell suspension. Cat No: 340242. 14. Borouitz M. J., Dave11 B. L., Boyett J. M., Fatella J. M., Jean J., Pullen D., Cdrist W. M., Bennet H. G. & Metzgar R. S. (1985) Monoclonal antibody definition of T-cell acute leukemia, A paediatric oncology group study. Blood

65, 785. 15. Poplak D. G. (1985) Acute lymphoblastic leukemia in childhood. Paediatr. clin. N. Am. 32, 669. 16. Sobol R. E., Royston L., Le Bien T. W., Minowada J., Anderson K. L., Davey F. R., Cultner J., Schiffer C., Elleson R. R. & Bloomfild C. D. (1985) Adult acute lymphoblastic leukemia phenotypes defined by monoclonal antibodies. Blood 65, 730. 17. Greves M. F. (1984) Subtypes of acute lymphoblastic

leukemia, Implications for the pathogenesis and epide-

124

K. R. Rajalekshmy et al.

miology of leukemia. In Puthogenesis of Leukemias and Lymphomas, Environmental Influences (Magrath I. T., O’Conor G. T. & Ramot B., Eds), p. 129. Raven Press, New York. 18. Kamat D. M., Gopal R., Advani S. H., Nair C. N., Kuman A., Saikia T., Nadkarni J. J. & Nadkrni J. S. (1985) Patterns of subtypes of acute lymphoblastic leukemia in India. Leukemia Rex 9, 927. 19. Bhargava M., Kumar R., Karak A., Kochupillai V., Arya L. S. & Mohana kumar T. (1988) Immunological subtypes of acute lymphoblastic leukemia in North India. Leukemia Res. 12, 8, 673678. 20. Kamel A. M., Assem M. M., Jaffer E. S., Magrath Moustala I., Dneen A. & Hindway D. S. (1989) Immunological phenotypica pattern of acute lymphoblastic leukemia in Egypt. Leukemia Res. 13, 519-525. 21. Ramot B. & Magrath I. T. (1982) Hypothesis, The environment is a major determinant of the immunological subtype of lymphoma and acute lymphoblastic leukemia in children. Br. J. Haemat. 50, 183. 22. Kasil E. G. (1990) Childhood leukemia. Is it a problem in tropical Africa? Leukemia Lymphoma 1, 187-193. 23. Greaves M. F. & Chan L. C. (1985) (Eds) Epidemiology of leukemia and lymphom. 24. Greaves M. F., Pegram S. M. & Chan L. C. (1985) Collaborative group study of the epidemiology of acute lymphoblastic leukemia subtypes. Background and first report. Leukemia Res. 9, 715-733. 25. Naeem S. & Hayee A. (1992) Acute lymphoblastic leukemia-a study of immunophenotypes. J. Pak. Med. Ass. 42, 83-86. 26. Fleming A. F. (1979) Epidemiology of the leukemias in Africa. Leukemia Res. 3, 51-60. 27. Greaves M. F., Raso J., Harire G., Verbi W., Catovsky D., Kung P. & Goldsteen G. (1981) Phenotypic heterogenicty

and cellular origins of T-cell malignancies. Leukemia Res. 5, 281. 28. Chirs W. M., Shuster J. J., Falletta J., Pullen D. J., Beard C. W. & Vetti T. J. (1988) Clinical features and outcome in childhood T-cell leukemia/lymphoma according to stage of lymphocytes differentiation. A paediatric oncology group study. Blood 72, 1891. 29. Kamel A. M., Ghaleb F. M., Assem M. M., Hindawy D. S., Jaffe E. S. & Magrath I. T. (1990) Phenotypic analysis of T cell acute lymphoblastic leukemia in Egypt. Leukemia Res. 14, 601. 30. Bernard A., Boumsell L., Reinherz E. L., Nadler L. M., Ritz I., Coppin H., Richard Y., Valensi F., Dauset J., Flandrin G., Lemerle J. & Lemerle S. F. (1981) Cell surface characterization of malignant T-cell from lymphoblastic lymphoma using monoclonal antibodies. Evidence for phenotypic differences between malignant T cells from patients with acute lymphoblastic leukemia and lymphoblastic lymphoma. Blood 57, 1105. 31. Cossman J., Chured M. M., Fisher R. L., Magrath I., Bollum F. & Jaffe E. S. (1983) Diversity of immunological phenotypes of lymphoblastic lymphoma. Cancer Res. 43, 4486. 32. Haddy T., Parker R. & Magrath I. T. (1989) Bone marrow involvement in young patients with Non-Hodgkin’s lymphoma, the importance of multiple bone marrow sample for accurate staging. Med. Puediut. Uncol. 17,418. 33. Pittaluga S., Raffeld M., Lipford E. H. & Crossman J. (1986) 3Al (cd7) expression precedes T, B gene rearrangement in precursor T lymphoblastic neoplasms. Blood 68, 134. 34. Haymes B. F. (1986) Summary of T cell studies performed deserving the second international workshop and conference on Human Leukocyte Differentiation Antigens. In Leukocyte Typing II (Reinherz E. L., Haynes B. F., Nadler L., Bernstein I. D., Eds), Vol. 1, p. 1. Springer, New York.