IMMUNOPATHOLOGY
Original Article
Identification of Monoclonal
Immunoglobulins and Quantitative
Immunoglobulin Abnormalities in Hairy Cell
Leukemia and Chronic Lymphocytic
Leukemia
DAVID A. HANSEN, MD, 1 BRUCE A. ROBBINS, MD, 1 DAVID J. BYLUND, MD, 1
LAWRENCE D. PIRO, MD, 2 ALAN SAVEN, MD, 2 AND DOUGLAS J. ELLISON, MD 1
Serum and urine samples from 161 cases of hairy cell leukemia (HCL)
and 50 cases of chronic lymphocytic leukemia (CLL) were analyzed for
monoclonal immunoglobulin (MIg) by using a combination of high-resolution protein electrophoresis, immunoelectrophoresis, and immunofixation. Quantitative immunoglobulin analysis also was performed on
all serum samples. Monoclonal immunoglobulin, usually of low intensity, was identified in serum or urine in 26 (16.1%) cases of HCL compared with 27 (54%) cases of CLL. Forty-eight (29.8%) cases of HCL
had an increase in one or more immunoglobulins; increases in IgG were
the most frequent. In CLL, 48 (96.0%) cases had a decrease in one or
more immunoglobulins, with decreases in IgG, IgA, and IgM in 76%,
68%, and 56% of the cases, respectively. The correlation between serum
or urine monoclonal immunoglobulin light chain and the surface membrane light chain was 88% in CLL compared with 47.4% in HCL. These
findings confirm previous observations of frequent polyclonal hyper-7globulinemia in HCL, hypo-7-globulinemia in CLL, and weak monoclonal immunoglobulins in both disorders. The contrasting immunoglobulin abnormalities in HCL and CLL indicate distinctive biologic
differences in these chronic B-cell leukemias. (Key words: Chronic lymphocytic leukemia; Hairy cell leukemia; Immunoglobulin quantitation;
Monoclonal immunoglobulins) Am J Clin Pathol 1994;102:580-585.
Hairy cell leukemia (HCL) is a rare chronic B-cell leukemia
characterized by pancytopenia, splenomegaly and the presence
of abnormal mononuclear cells with characteristic cytoplasmic
projections in the blood, bone marrow and spleen.1"3 Previous
studies have shown a polyclonal increase in immunoglobulins
in 18% to 30% of the cases3"5; increases in IgG were the most
common. 6 Hypo-7-globulinemia is not a common finding in
HCL. Monoclonal gammopathy has been reported to be rare,
with the incidence ranging from < 1% to 8% using protein electrophoresis and immunoelectrophoresis (IEP).1'3-6"10 Although
several cases of HCL with monoclonal proteins have been reported, synthesis of the monoclonal immunoglobulin (MIg) by
the hairy cells has not been definitely established. 1'7"9""15 In
some cases, multiple myeloma was thought to coexist with
HCL. 9 ' 3 1 4
In contrast to HCL, hypo-7-globulinemia is frequently
noted in chronic lymphocytic leukemia (CLL) and has been
reported to occur in 10% to 75% of cases.16"20 The degree of
hypo-7-globulinemia correlates with clinical stage, increasing
in severity as stage increases.' 617 Polyclonal hyper-7-globulinemia is rarely reported in CLL. The frequency of identification
of MIg in CLL has been reported to range from 5% to 74%
depending on whether IEP or immunofixation (IFIX) was used
as the detection method.21"26
In this study, serum and urine samples from 161 cases of
HCL and 50 cases of CLL were analyzed for MIg using a combination of high-resolution protein electrophoresis, IEP and
IFIX. Quantitative immunoglobulin analysis was further performed on all serum samples. The intent of this study was to
document and compare the frequency of immunoglobulin abnormalities in a large series of HCL and CLL. By further comparison with membrane phenotypes, these data may provide
insights into the mechanisms responsible for MIg production
and the nature of immune dysregulation in these related leukemias.
MATERIALS A N D METHODS
From the 'Department ofPathology and the2Division ofHematology
and Medical Oncology, Scripps Clinic and Research Foundation, La Case Specimens
Jolla. California.
The specimens from all patients with HCL (age range, 30-79
years; median age, 56 years) were obtained while the patients
Manuscript received August 9, 1993; revision accepted October 25,
were being evaluated at Scripps Clinic and Research Founda1993.
tion (La Jolla, CA) between May 1, 1990, and August 1, 1992,
Address reprint requests to Dr. Robbins: Scripps Clinic and Refor treatment with 2-chlorodeoxyadenosine. Standard morphosearch Foundation, Department of Pathology, 211C, 10666 North
logic criteria including examination of bone marrow, periphTorrey Pines Road, La Jolla, CA 92037.
580
HANSEN ET AL.
Immunoglobulins
in Hairy Cell Leukemia
eral blood, and tartrate-resistant acid phosphatase stains, and
membrane phenotyping were used to confirm all diagnoses of
HCL. Morphologic variants of HCL were excluded from the
study. The HCL patients were either previously untreated or
had undergone previous splenectomy or chemotherapy with
a-interferon or 2-deoxycoformycin. Standard morphologic criteria and membrane phenotyping were employed to confirm
the diagnoses in all CLL patients. All CLL patients (age range,
43-87 years; median age, 67 years) were of Rai clinical stage 3
or4and had received previousalkylator-containingchemotherapy.
High Resolution
Protein
Electrophoresis
Serum and urine samples were tested using high resolution
agarose protein electrophoresis (Titan Gel High Resolution
Protein System, Helena Laboratories, Beaumont, TX). Urine
samples were concentrated 100X before electrophoresis (Amicon, Beverly, MA).
TABLE 1. MONOCLONAL IMMUNOGLOBULINS
HCL
n
Serum
IgG
Total
Kappa
Lambda
IgA
Total
Kappa
Lambda
IgM
Total
Kappa
Lambda
Total
161
16(9.9)
10(6.2)
6 (3.7)
Urine
Kappa light chains
Lambda light chains
Total
CLL
50
11(22.0)
10(20.0)
1 (2.0)
2(1.2)
1 (0.6)
1 (0.6)
0(0)
0 (0)
0 (0)
1(0.6)
0(0)
1 (0.6)
19(11.8)
16(32.0)
12(24.0)
4 (8.0)
27(54.0)
7(4.4)
2(1.2)
9(5.6)
12(24.0)
4(8.0)
16(32.0)
Values arc no. (%).
Immunoelectrophoresis
Serum and urine samples were tested using the IEP procedure provided by Gelman Instrument (Ann Arbor, MI). Initially, each serum specimen was tested using anti-whole human serum (Atlantic Antibody, Stillwater, MN) and
anti-polyvalent human immunoglobulin (Whittaker, Walkersville, MD). If the initial results of IEP were abnormal, the
specimen was studied with monospecific antiserum for IgG,
IgA, IgM (Atlantic Antibody, Stillwater, MN), anti-x, and antiX (Whittaker, Walkersville, MD). Initial screening of urine specimens included anti-whole human serum (Atlantic Antibody),
and anti-polyvalent human immunoglobulins (Whittaker;
DAK.O, Carpinteria, CA). Urine specimens containing identifiable immunoglobulins were further evaluated in a similar
fashion to serum with the addition of antibodies specific for
human free K and X light-chain determinants (Kallstead,
Chaska, MN).
Immunofixation
Serum specimens with weak restricted bands or oligoclonal
bands on high-resolution protein electrophoresis were more
sensitively characterized by IFIX than IEP. Electrophoresis
was carried out using an agarose high resolution protein system
(470682, Ciba Corning Diagnostics, Palo Alto, CA). The serum
samples were electrophoresed at 230 V for 20 minutes. Monospecific antisera for IgG, IgA and IgM (Ciba Corning Diagnostics) were used. Anti-K and anti-X light-chain reagents (Whittaker) were each applied at two dilutions to optimize reactivity.
The gels were stained with acid violet after washing to remove
nonprecipitated protein.
Immunoglobulin
581
Quantitation
Quantitation of serum IgG, IgA, and IgM was performed
using rate nephelometry (Array Protein System, Beckman Instruments, Brea, CA). Reference ranges included: IgG (7231685 mg/dL), IgA (69-382 mg/dL), and IgM (63-277 mg/dL).
Vol. I
Membrane
Phenotypes
The immunophenotypic features of the cells from these
HCL and CLL patients have been reported in a previous
study.27 Peripheral blood mononuclear cells from all HCL and
CLL cases were isolated and the membrane phenotypes were
analyzed using two-color flow cytometry as described in detail
in the previous study.27
RESULTS
Serum MIg were identified in 19(11.8%) of 161 cases of
HCL studied. Of these, monoclonal IgG proteins (84.2%) constituted the majority as demonstrated in Table 1 and most were
weak in intensity. In cases with serum monoclonal IgG proteins, K light chains predominated. Urine monoclonal free light
chains were detected in 9 (5.6%) cases with K light chains again
predominating. Monoclonal immunoglobulin in either serum
or urine were found in 26 (16.1%) cases.
In the 50 cases of CLL studied, 27 (54%) had a detectable
MIg. IgM comprised 16 (59.3%) of 27 and the remainder were
IgG, as shown in Table 1. K Light chains were present in 22
(81.5%) of 27 cases and represented a higher percentage than
that seen in HCL. Urine monoclonal free light chains were
detected in 16 (32%) of the CLL cases; 12 of these were K light
chains. All cases with a urine monoclonal free light chain also
had a serum MIg.
Correlation between the light chain of the MIg and the surface membrane light chain is shown in Table 2. The correlation
between the light chains in CLL is 88%, whereas the correlation
between HCL surface membrane light chain and MIg light
chains is only 47.4%. In CLL, all urine MIg light chains were
identical to the surface membrane phenotype light chain,
whereas in HCL the urine MIg light chain corresponded to the
surface membrane light chain in 75% of the cases.
Forty-eight (29.8%) of 161 cases of HCL had an increase in
one or more immunoglobulins (Table 3) (Fig. 1) with increases
of IgG the most frequent, as seen in 35 (21.7%) cases. Increases
• No. 5
582
IMMUNOPATHOLOGY
Original Article
TABLE 2. CORRELATION OF SERUM/URINE MIg LIGHT
CHAIN WITH SURFACE MEMBRANE LIGHT CHAIN
HCL
CLL
Serum
IgG
IgA
IgM
Total
7/16(43.8)
2/2 (100)
0/1 (0)
9/19(47.4)
9/9 (100)*
0(NA)
13/16(81.0)
22/25 (88.0)
Urine
Kappa light chains
Lambda light chains
Total
4/6 (66.7)t
2/2 (100)
6/8 (75.0)
12/12(100)
4/4 (100)
16/16(100)
Values arc no. of cases wilh light chain identity (%).
* Two cases were not included because the surface membrane phenotype was not determined.
t One case were not included because the surface membrane phenotype was not determined.
NA = Not Applicable.
in IgA and IgM were seen in 12 (7.4%) and 13 (8.0%) cases,
respectively. Elevation of two immunoglobulins was noted in 8
(5.0%) cases and increases of all three immunoglobulins was
present in 3 (1.9%) cases. More than 57% of all HCL cases had
no quantitative immunoglobulin abnormality.
Twenty-eight (17.4%) of the HCL cases showed a decrease of
one or more immunoglobulins. Decreases in IgM were the
most frequent, as demonstrated in 25 (15.5%) cases. Decreases
in IgG and IgA were relatively rare and were present in 4 (2.5%)
and 2 (1.2%) cases, respectively. Only 3 (1.9%) of cases showed
a decrease of two immunoglobulins, and no cases had a decrease in all three immunoglobulins.
In contrast to HCL, 48 (96%) of the 50 CLL cases had a
decrease in one or more immunoglobulins. IgG was the immunoglobulin most frequently decreased as shown in 38 (76%)
cases. IgA and IgM were decreased in 34 (68%) and 28 (56%) of
cases, respectively. Decreases in two or more immunoglobulins
were quite common, with 36 (72%) cases having two immunoglobulins decreased and 16 (32%) cases having all immunoglobulins decreased.
Only 4 (8%) cases of CLL showed an increase of one or more
immunoglobulins. In all four cases, this represented an in-
crease in the IgM fraction. These four cases had weak to moderate levels of a monoclonal IgM protein, accounting for the
increases seen. Otherwise, increases in serum immunoglobulins were not demonstrated.
DISCUSSION
Monoclonal gammopathy in HCL has been identified in a
few case reports.1,7"9'""15 In our study 16% of 161 cases of HCL
had a MIg in the serum or urine identified using high-resolution protein electrophoresis, IEP, and IFIX. This percentage is
twice that of the highest value previously reported1'3,6"'0 and
may be the result of using sensitive techniques such as high-resolution protein electrophoresis and IFIX instead of only IEP.
The majority of the MIg in our study were of weak intensity
and could be overlooked if only IEP techniques were used.
Approximately 84% of the serum MIg were IgG, similar to
previous data."
Monoclonal gammopathy in CLL has been reported to
range from 5% to 74% depending on the sensitivity of the test
procedures used.21"26 Our finding of MIg in 54% of CLL cases is
consistent with more recent reports using sensitive methods.21'22'24 Previous studies identified monoclonal IgG as the
most common isotype,24'26 but in our study IgM was identified
more frequently than IgG. In CLL 88% of the serum or urine
MIg light chains were of the same class as the surface membrane light chain. These data are consistent with previous studies using idiotype analysis that have suggested that in most
cases, the serum or urine MIg appears to originate from the
CLL B cells.28
In our study, only 47% of the HCL cases with MIg had identical light chains on the MIg and the hairy cell surface membrane. This lack of light-chain correlation suggests that in
many cases the MIg may not be produced by the HCL clone.
Studies of healthy populations using IFIX techniques have
demonstrated serum MIgs in 5% to 6% of individuals in the age
range of our HCL patients.29,30 The occurrence of serum MIgs
in the general population could account for up to half of the
11.8% incidence of serum MIgs detected in HCL. Some case
reports of MIg in HCL have described a coexisting second lymphoid disorder, including multiple myeloma. 91314 Morphologic examination of the bone marrow in HCL cases with MIg
TABLE 3. IMMUNOGLOBULIN QUANTITATION
HCL
50
161
IgG* Mean ± SD (mg/dL)
Increased
Normal
Decreased
IgA* Mean ± SD (mg/dL)
Increased
Normal
Decreased
IgM* Mean ± SD (mg/dL)
Increased
Normal
Decreased
P Value
CLL
<0.001
587 ±218
1,469 ±697
0(0)
12(24.0)
38 (76.0)
35(21.7)
121 (75.2)
5(3.1)
<0.001
61 ± 42
219 ± 114
0(0)
16(32.0)
34 (68.0)
12(7.5)
147(91.3)
2(1.2)
101 ± 118
148+ 182
13(8.1)
123(76.4)
25(15.5)
Values are no. (%).
* Reference ranges: IgG: 723-1,685 mg/dL: IgA: 69-382 mg/dL: IgM: 63-277 mg/dL.
A.J.C.P. • November 1994
<0.091
4 (8.0)
18(36.0)
28 (56.0)
HANSEN ET AL.
Immunoglobulins
583
in Hairy Cell Leukemia
HCL
CLL
D
20
15-
10
0
4*1
500
1000
1500
llMMllll
II M
2000 2500 3000 3500
4000+
0+-r
0
j|
500
im.
1000
IgG (mg/dl)
1500
2000
2500
3000
3500
4000+
IgG (mg/dl)
HCL
CLL
20
15-
10
0
100
200
300
400
500
600
700+
IgA (mg/dl)
200
300
400
500
600
700+
500
600
700+
IgA (mg/dl)
HCL
C
25.
ll
04-*#*#
0
100
20
CLL
15
10-
•••ill
0
100
200
300
400
I 100
M I I I200
M , 300
I
•
500
600
700 +
400
IgM (mg/dl)
IgM (mg/dl)
FIG. 1. Immunoglobulin levels in HCL and CLL. A, IgG HCL; B, IgA HCL; C, IgM HCL; D, IgG CLL; E. IgA CLL; and F, IgM CLL
in our study failed to demonstrate evidence of a second lymphoproliferative disorder.
Nearly 30% of the HCL cases had a quantitative increase of
IgG, IgA or IgM, similar to previous reports.3"6 There has been
very little investigation into the pathogenesis of polyclonal
hyper-7-globulinemia in HCL. One recent study suggests that
HCL patients with persistent polyclonal hyper-7-globulinemia
may secrete a soluble factor which, in the presence of T cells,
induces activation of normal B cells followed by increased production of IgG. No such IgG production-enhancing activity
was isolated from HCL patients without persistent polyclonal
hyper-7-globulinemia.31
Hypo-7-globulinemia in CLL is a frequent and well documented phenomenon 16 " 20 and is present at the time of diagnosis in about 20% of the cases.16,17 In our study, 96% of the cases
had a decrease in at least one immunoglobulin. The frequency
Vol. 102-No. 5
584
IMMUNOPATHOLOGY
Original Article
of decreased IgG and the number of cases with two or more
immunoglobulins decreased in our study is greater than previously reported.18 Previous reports have shown the frequency
and magnitude of the hypo-7-globulinemia in CLL increases
with duration of disease and advanced clinical stage. 16 ' 7 ' 9,32
Chemotherapy has also been shown to aggravate hypo-7globulinemia.16 All of our CLL patients were in Rai stage 3 or 4
and the advanced clinical stage and previous chemotherapy
probably contributed to the high percentage of patients with
hypo-7-globulinemia.
Various factors have been described to account for the decreased immunoglobulin synthesis in CLL including impaired
B-cell function,20,33 a T-cell regulatory abnormality involving T
helper cells, and large granular lymphocytes from patients with
B-cell CLL suppressing normal B cells.33"35 An additional possible mechanism involves transforming growth factor /3 (TGF(8), which is a negative autocrine regulator of normal B-lymphocyte growth and differentiation36"38 that inhibits B-cell
growth factor-induced proliferation of B lymphocytes.39
Chronic lymphocytic leukemia B cells have been shown to be
resistant to TGF-/338 and can also express and release TGF-18.37
By these mechanisms, the CLL B cell could escape from negative regulatory control while at the same time causing immunoglobulin production by normal B cells to be decreased. Hairy
cell leukemia has been shown to be resistant to TGF-/J but has
not been shown to express or release TGF-/S.34 Whether this
lack of expression or secretion of TGF-/3 is related to the
changes in immunoglobulin levels in HCL is unknown.
Immunophenotyping of HCL has shown that hairy cells are
usually positive for surface membrane immunoglobulin,
CD 19, CD20, CD1 lc, CD25, CD 103 (B-ly7), and PCA-1, and
negative for CD5, CD21, and PC-1. 40 " 42 B cells in CLL are
typically positive for surface membrane immunoglobulin,
CD 19, CD20, CD21, and CD5. 40,43 The hairy cell surface
membrane immunoglobulin frequently contains IgG, and simultaneous expression of multiple heavy-chain isotypes has
been observed.41 In CLL the surface membrane immunoglobulin is predominantly IgM ± IgD and rarely IgG or IgA.40 The
frequent presence of surface membrane IgG and IgA, expression of PCA-1, and the absence of CD21 and PC-1 suggests that
the hairy cell may correspond to a late stage of B-cell differentiation prior to becoming a plasma cell.5,42 The immunophenotype of B-cell CLL corresponds to the CD5-positive B-cell subset, which is at an intermediate stage of differentiation.43,44 The
frequent presence of serum IgG MIgs in CLL suggests that the
CLL B cell may have the capability of further differentiation
and heavy-chain class switching. Considering that HCL has
been considered a "preplasma cell," we were interested to observe a higher frequency of serum and urine MIgs in CLL.
Although we have documented quantitative immunoglobulin abnormalities and the presence of MIg in HCL, the cause of
these abnormalities is not understood. Although MIgs in CLL
have been shown to be related to surface membrane immunoglobulin by idiotype analysis,28 the same study has not been
performed with MIg in HCL. Additional investigations are
needed to define the origin of MIgs in HCL.
Acknowledgments. The authors appreciate the excellent technical assistance provided by Sharon Gehrke, Cristin Carey, Mila Fontimajor,
Jeannette Diaz, Laura Calafata, Nancy Martin, and Thelma Convento.
We thank Dr. James Koziol for assistance with statistical analysis. We
also thank Marcia Filbert, Julie Koehler and Linda Miller for their
outstanding word processing services.
REFERENCES
1. Westbrook CA, Golde DW. Clinical problems in hairy cell leukemia: Diagnosis and management. Semin Oncol 1984; 11:514522.
2. Westbrook CA, Golomb HM. Hairy-cell leukemia. Curr Concepts
Oncol 1984;Winter:4-9.
3. Flandrin G, Sigaux F, Sebahoun G, Bouffette P. Hairy cell leukemia: Clinical presentation and follow-up of 211 patients. Semin
Oncol 1984; 11:458-471.
4. Golomb HM, Catovsky D, Golde DW. Hairy cell leukemia: A
clinical review based on 71 cases. Ann Intern Med 1978,89:677683.
5. Paoletti M, Bitter MA, Vardiman JW. Hairy cell leukemia: Morphologic, cytochemical, and immunologic features. Clin Lab
Med 1988;8:179-195.
6. Farcet J-P, Weschsler J, Wirquin V, et al. Vasculitis in hairy-cell
leukemia. Arch Intern Med 1987; 147:660-664.
7. Cohen O, Mor F, Beigel Y, et al. The significance of paraproteinemia in hairy cell leukemia: Case report and review of the literature. Haematologica 1990; 75:179-181.
8. Damasio EE, Spriano M, Repetto M, et al. Hairy cell leukemia: A
retrospective study of 235 cases by the Italian Cooperative
Group (ICGHCL) according to Jansen's clinical staging system.
Acta Haematol 1984;72:326-334.
9. Catovsky D. Costello C. Loukopoulos D, et al. Hairy cell leukemia
and myelomatosis: Chance association or clinical manifestations of the same B-cell disease spectrum. Blood 1981; 57:758763.
10. Jansen J, Bolhuis RLH. van Nieuwkoop JA, et al. Paraproteinaemia plus osteolytic lesions in typical hairy-cell leukemia. Br J
Haematol 1983;54:531-541.
11. Cawley JC, Burns GF, Bevan A. et al. Typical hairy-cell leukaemia
with IgGk paraproteinaemia. Br J Haematol 1979;43:215-221.
12. Giardina SL. SchrofT RW, Woodhouse CS, et al. Detection of two
distinct malignant B cell clones in a single patient using antiidiotype monoclonal antibodies and immunoglobulin gene rearrangement. Blood 1985;66:1017-1021.
13. Lawlor E, Willoughby R, O'Briain DS. Daly PA. Multiple neoplasms in hairy cell leukaemia. Acta Haematol 1984;72:57-60.
14. Noseda G, Reiner M, Scali G. Haarzell-leukamie und myelom.
Schweiz Med Wochenschr 1980; 110:1494-1498.
15. Golde DW, Saxon A, Stevens RH. Macroglobulinemia and hairycell leukemia. N EnglJ Med 1977;296:92-93.
16. Itala M, Helenius H, Nikoskelainen J, Remes K. Infections and
serum IgG levels in patients with chronic lymphocytic leukemia. Eur J Haematol 1992;48:266-270.
17. Ben-Bassat I, Many A, Modan M, et al. Serum immunoglobulins
in chronic lymphocytic leukemia. Am J Med Sci 1979;278:4-9.
18. Rozman C, Montserrat E, Vinolas N. Serum immunoglobulins in
B-chronic lymphocytic leukemia: Natural history and prognostic significance. Cancer 1988;61:279-283.
19. Orfao A, Gonzalez M.San Miguel J F.etal. Surface phenotype and
immunoglobulin levels in B-cell chronic lymphocytic leukaemia. Haematologica 1990;23:49-56.
20. Dighiero G. An attempt to explain disordered immunity and hypogammaglobulinemia in B-CLL. Nouv Rev Fr Hemalol
1988;30:283-288.
21. Katsikis PD, Pavlidis NA, Papadopoulos NM, et al. Monoclonal
immunoglobulins and autoantibodies in lymphoid malignancies. Anticancer Res 1990; 10:1719-1724.
22. Deegan MJ, Abraham JP, Sawdyk M, Van Slyck EJ. High incidence of monoclonal proteins in the serum and urine of chronic
lymphocytic leukemia patients. Blood 1984;64:1207-1211.
23. Pezzoli A, Pascali E. Monoclonal proteins in chronic lymphocytic
leukemia. Am J Clin Pathol 1987;88:543. Letter.
24. Pezzoli A, Pascali E. Monoclonal Bence Jones proteinuria in
chronic lymphocytic leukaemia. Scand J Haematol 1986;
36:18-24.
25. Stevenson FK, Spellerberg M, Smith JL. Monoclonal immunoglobulin light chain in urine of patients with B lymphocytic
A.J.C.P. • November 1994
HANSEN ET AL.
Immunoglobulins in Hairy Cell Leukemia
26.
27.
28.
29.
30.
31.
32.
33.
34.
disease: Its source and use as a diagnostic aid. Br J Cancer
1983;47:607-612.
Noel P, Kyle RA. Monoclonal proteins in chronic lymphocytic
leukemia. Am J Clin Pathol 1987:87:385-388.
Robbins BA, Ellison DJ. Spinosa JC, et al. Diagnostic application
of two-color flow cytometry in 161 cases of hairy cell leukemia.
Blood 1993:82:1277-1287.
Qian G-X. Fu SM. Solanki DL. Rai KR. Circulating monoclonal
IgM proteins in B cell chronic lymphocytic leukemia: Their
identification, characterization and relationship to membrane
IgM. J Immunol 1984:133:3396-3400.
Papadopoulos NM. Elin RJ. Wilson DM. Incidence of gammaglobulin in a healthy population by high-resolution electrophoresis. Clin Clwm 1982:28:707-708.
Crawford J, Eye MK. Cohen HJ. Evaluation of monoclonal gammopathies in the 'well' elderly. Am J Med 1987;82:39-45.
Nakamura Y. Machii T, Tokumine Y. et al. Hairy cells from hairy
cell leukemia patients presenting with pronounced polyclonal
hypergammaglobulinemia secrete a factor enhancing IgG synthesis. Clin Immunol Immunopathol 1993;66:212-218.
FiddesP. Penny R.Wells J V. Rozenberg MC. Clinical correlations
with immunoglobulin levels in chronic lymphatic leukaemia.
Ausi NZJ Med 1972:4:346-350.
Foon KA. Rai KR. Gale RP. Chronic lymphocytic leukemia: New
insights into biology and therapy. Ann Intern Med
1990:113:525-539.
Kay NE. Perri RT. Evidence that large granular lymphocytes from
B-CLL patients with hypogammaglobulinemia down-regulate
B-cell immunoglobulin synthesis. Blood 1989;73:1016-1019.
585
35. Zaknoen SL, Kay NE. Immunoregulatory cell dysfunction in
chronic B-cell leukemias. Blood Rev 1990;4:165-174.
36. Kremer J-P, Reisbach G, Nerl C, Dormer P. B-cell chronic lymphocytic leukaemia cells express and release transforming growth
factor-/?. Br J Haematol 1992;80:480-487.
37. Kehrl JH, Roberts AB. Wakefield LM, et al. Transforming growth
factor 8 is an important immunomodulatory protein for human
B lymphocytes. J Immunol 1986; 137:3855-3860.
38. Keller JR, SingGK, Ellingsworth LR, Ruscetti FW. Transforming
growth factor B: Possible roles in the regulation of normal and
leukemic hematopoietic cell growth. J Cell Biochem
1989;39:175-184.
39. Peiti-Koskas E, Genot E, Lawrence D, Kolb J-P. Inhibition of the
proliferative response of human B lymphocytes to B cell growth
factor by transforming growth factor-beta. Eur J Immunol
1988;18:111-116.
40. Golomb HM, Davis S, Wilson C, Vardiman J. Surface immunoglobulins on hairy cells of 55 patients with hairy cell leukemia.
Am J Hematol 1982;12:397-401.
41. Dalai BI, Fitzpatrick LA. Hairy cell leukemia: An update. Lab
Med 1991;22:31-36.
42. Hassan IB. Hagberg H, Sundstrom C. Immunophenotype of hairy
cell leukemia. Eur J Haematol 1990;45:172-176.
43. Dighiero G, Travade P, Chevret S, et al. B-Cell chronic lymphocytic leukemia: Present status and future directions. Blood
1991:78:1901-1914.
44. Kay NE. Perri RT. Immunobiology of malignant B cells and immunoregulatory cells in B-chronic lymphocytic leukemia. Clin
Lab Med 1988;8:163-177.
Vol. 102-No. 5