Human monoclonal macroglobulins with antibody activity

Human Monoclonal Macroglobulins With Antibody Activity Marvin J. Stone, Yolonda G. McElroy, Alan Pestronk, Janet L. Reynolds, Joseph T. Newman, and Alex W. Tong Assays for specific antigen-binding activity were performed on sera from 172 patients with monoclonal macroglobulinemia defined by immunofixation electrophoresis. The sera were collected between 1970 and 2002. Mean IgM level was 1,409 mg/dL with a range from 70 to 6,800. Cryoglobulins were identified in 15.3% (26/170 sera: 12 trace, five single component, and nine mixed IgM-IgG). Rheumatoid factor (RF) was detected in 19 of 151 (12.6%) samples with titers ranging from 1:80 to 1:327,680. Among the nine mixed IgM-IgG cryos, eight were RF-positive and six of six displayed positivity for hepatitis C virus. Cold agglutinins (CA) were present in 8.5% (10/117) of sera with anti-I titers between 1:512 and 1:65,536. IgM binding to a series of glycosaminoglycan oligosaccharides, glycolipids, and glycoprotein antigens was found in 75 samples (43%). IgM binding to antigens having known associations to polyneuropathies occurred in 20 patients (12%). Antinuclear antibody (ANA) was documented in 10.7% (18/ 169) of sera. Anti-DNA activity was absent in all samples tested. Sera from 71% of patients with monoclonal macroglobulinemia in this series exhibited binding to autoantigens. Some of these immune complexes resulted in clinically significant manifestations. Our results suggest that many monoclonal immunoglobulins may be functional antibodies rather than “paraproteins.” Characterization of antigen-binding activities may provide insight into the pathogenesis of monoclonal gammopathies. Semin Oncol 30:318-324. © 2003 Elsevier Inc. All rights reserved.

M

ONOCLONAL macroglobulins, myeloma immunoglobulins, and Bence Jones proteins from patients and mice have played a central role in elucidation of normal immunoglobulin structure, genetics, synthesis, and metabolism.1 Most of these monoclonal proteins have not shown evident antibody activity. Specific antigen binding occasionally had been associated with Waldenstrom macroglobulins as described by the Dacie,

From the Baylor Charles A. Sammons Cancer Center and Immunology Laboratory, Baylor University Medical Center, Dallas, TX; and the Department of Neurology, Washington University School of Medicine, St Louis, MO. Supported in part by the Edward and Ruth Wilkof Foundation, the Robert Schanbaum Memorial Fund, and the James E. Nauss Cancer Research Fund. Address reprint requests to Marvin J. Stone, MD, Baylor-Sammons Cancer Center, 3500 Gaston Ave, Dallas, TX 75246. © 2003 Elsevier Inc. All rights reserved. 0093-7754/03/3002-0042$30.00/0 doi:10.1053/sonc.2003.50077 318

Kunkel, and Waldenstro¨m groups during the period from 1957 to 1964.2-4 The activities noted were cold agglutinins and rheumatoid factors. During the next decade, studies on two Waldenstrom macroglobulins with anti-IgG activity (mixed cryoglobulins) demonstrated that these monoclonal IgMs satisfied operational criteria for functional antibodies (Table 1).5-11 Subsequently, some other Waldenstrom macroglobulins and human myeloma proteins have been shown to have antibody activity, principally to autoantigens and bacterial antigens. The purpose of this study was to identify antibody activity in 172 sera from patients with monoclonal macroglobulinemia collected over a 32-year interval. MATERIALS AND METHODS Analysis was carried out on sera from 172 patients with monoclonal macroglobulinemia collected between 1970 to 2002 (12 from 1970 to 1979, 17 from 1980 to 1989, 53 from 1990 to 1995, 75 from 1996 to 2000, nine in 2001, and six in 2002). Sera were stored at ⫺20°C until 1 to 20 days before assay. Clinical data were determined by retrospective chart review. The study protocol was approved by the Institutional Review Board, Baylor University Medical Center, Dallas, TX.

Immunofixation Electrophoresis, Serum Protein Electrophoresis, and Immunoglobulin Quantification Monoclonal macroglobulinemia was determined by serum immunofixation electrophoresis (IFE) showing a homogeneous band against monospecific IgM antiserum and a single light (L)-chain type (kappa or lambda), with the Paragon system (Beckman Coulter, Fullerton, CA).12 Serum protein electrophoresis (SPE) was performed with the SPIFE 2000 automated analyzer (Beckman Coulter).13 The normal ranges (g/dL) in healthy adults are: total protein, 6.0 to 8.5; albumin, 3.2 to 5.5; alpha-1, 0.1 to 0.4; alpha-2, 0.6 to 1.0; beta, 0.7 to 1.1; and gamma, 0.8 to 1.6. An immunoturbidimetric assay was used to quantify serum IgM, IgG, or IgA.14 The normal values in adult serum are 44 to 215 mg/dL for IgM, 534 to 1,326 mg/dL for IgG, and 30 to 318 mg/dL for IgA. Relative serum viscosity was determined as previously described.11 Identification and characterization of cryoglobulinemia. Cryoglobulins (cryo) are immunoglobulins which precipitate or gel at temperatures ⬍37°C and dissolve on rewarming.10,11,15 Cryoglobulin components in redissolved cryoprecipitates were determined as described previously.10,11

Rheumatoid Factor and Cold Agglutination Determinations Rheumatoid factor (RF) in patient sera was identified by agglutination with the Fc region of heat-denatured human IgG Seminars in Oncology, Vol 30, No 2 (April), 2003: pp 318-324

HUMAN MONOCLONAL MACROGLOBULIN ANTIBODIES

319

Table 1. Properties of Two Waldenstrom’s Macroglobulin Antibodies With Anti-IgG Activity Property

IgMLay

IgMSie

Euglobulin RF activity Binds native (monomer) IgG KA for IgG1, K Thermodynamic parameters ⌬F° (kcal 䡠 mol⫺1) ⌬H° (kcal 䡠 mol⫺1) ⌬S° (cal 䡠 deg⫺1 䡠 mol⫺1) Antigenic specificity Antigen-antibody complex Optimal pH for precipitation of complex Valence Pentamer IgM Isolated Fab␮ Effective valence of antigen Cross-reactivity Lorisiforme IgG Rabbit IgG V region subgroups L chain H chain Cross-idiotypic specificity Complex fixes complement

No Yes Yes 6.8 ⫻ 104 M⫺1

No Yes Yes 4.5 ⫻ 104 M⫺1

⫺5.8 ⫾ 0.2 ⫺7 ⫾ 1 ⫺5 ⫾ 1 Fc␥ Cryoglobulin 6.0

⫺5.7 ⫾ 0.1 ⫺4.9 ⫾ 1.0 2.7 ⫾ 0.3 Fc␥ Cryoglobulin 7.4

5 10 1

5 to 10 10 1

Yes No

Not tested Yes

V KI VHIII Po system No

VKIIIb Not known Wa system No

NOTE. Data from references 1 and 11.

coated onto the surface of biologically inert latex particles.16 A positive reaction was defined as a serial dilution greater than 1:80. Cold agglutinin (CA) reactivity against the blood group I or i antigens was determined after incubation with 2% human type O red blood cells at various temperatures.17 A titer of ⱖ1:64 was considered positive.

Reactivity to Neural Antigens Enzyme-linked immunosorbent (ELISA) assays measured IgM binding to 16 purified antigens, including glycolipids, oligosaccharides, proteins, and glycoproteins.18 Western blotting examined IgM binding to proteins in myelin, and neuronal nuclei and cytoplasm.19

Antinuclear Antibody and Anti-DNA Activity Serum antinuclear antibody (ANA) activity was detected by an indirect immunofluorescence test with human epithelial Hep-2 cells transfected with specific DNA sequence for the SS-A (Ro) autoantigen.20 A screening titer of greater than 1:40 was considered positive. Serum anti-DNA antibody reactivity was determined by an indirect immunofluorescent antibody test that utilized the trypanosome, Crithidia luciliae as substrate.21

Hepatitis C Virus Antibody and RNA Anti– hepatitis C virus (HCV) serum antibody was detected by the Vitros ECi assay (Ortho-Clinical Diagnostics, Raritan, NJ). Serum detection of HCV RNA was performed by a qual-

itative assay with a sensitivity of ⱖ5 IU/mL.22 HCV RNA was quantified by the branched DNA assay with a range of 615 IU/mL to 7,700,000 IU/mL (1 IU ⫽ 5.2 copies).23

RESULTS

Mean age of the patients was 67 years (range, 31 to 96). There were 90 men and 82 women. Of the original 172 patient sera, one could not be verified as containing monoclonal macroglobulin and was excluded. The remaining 171 had monoclonal macroglobulins by IFE. Quantitative serum IgM levels varied between 70 and 6,800 mg/dL, with a mean level of 1,409. Some patients did not have a definable M-protein on SPE because of polyclonal hypergammaglobulinemia or low serum IgM levels. Biclonal gammopathies were observed in nine patients. Cryoglobulins and RF Twenty-six patients had cryoglobulins (12 trace positive, five single component, and nine mixed IgM-IgG) (Table 2). The previously described Waldenstrom macroglobulin antibody (IgMSie) with anti-IgG activity1,10,11 retained its cryopre-

320

STONE ET AL

Table 2. Cryoglobulin Patients Patient

Year

M-Spike

L Chain

Cryoglobulin

RF

HCV Antibodies

HCV RNA*

1CG 2CG 3CG 4CG 5CG 6CG 7CG 8CG 9CG 10CG 11CG 12CG 13CG 14CG 15CG 16CG 17CG 18CG 19CG 20CG 21CG 22CG 23CG 24CG 25CG 26CG

1985 1992 1994 1995 1995 1997 1999 2000 2000 1993 1991 1998 1996 1981 1996 2001 1978 1982 1995 1983 1995 2001 1970 1997 2000 1979

1.7 1.5 None Hypo 0.6 2.7 PHG, 3.1 Hypo 0.6 None 2.0 0.4 2.1 3.5 3.5 4.3 5.5 4.1 PHG, 1.0 1.5 4.7 4.7 2.8 3.1, 0.2 0.4 0.9

Kappa Kappa Lambda Kappa Kappa Kappa Lambda Kappa Lambda Kappa Kappa Lambda Kappa Kappa Kappa Kappa Lambda Kappa Kappa Kappa Kappa Kappa Kappa Kappa Kappa Lambda

Trace Trace Trace Trace Trace Trace Trace Trace Trace Trace Trace Trace Single Single Single Single Single Mixed Mixed Mixed Mixed Mixed Mixed Mixed Mixed Mixed

(⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) NT (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) 1:5120 (⫹) 1:81,920 1:327,680 1:256,000 1:10,240 1:640 (⫹)

neg pos NT neg neg neg neg neg NT neg neg NT neg neg NT neg neg neg NT pos pos pos pos pos pos NT

neg 60,277 NT neg neg neg neg ⬎5 NT neg neg NT neg neg NT neg ⬎5 neg NT ⬎5 677 30,702 ⬎5 123,205 616,710 NT

Abbreviations: PHG, polyclonal hypergammaglobulinemia; NT, not tested; neg, negative; pos, positive; Hypo, hypogamma globulinemia. * HCV RNA in IU/mL; 1 IU ⫽ 5.2 copies.

cipitating property with antigen more than 32 years after the serum was collected (Fig 1). Of the cryoglobulin-positive patients, six had lambda L chains. Two of 26 cryoglobulin patients’ sera were tested for HCV. One trace positive cryoglobulin patient also was positive for hepatitis C antibody and RNA, although RF was negative. Thirteen patients with cryoglobulinemia were negative for hepatitis C antibody; two of these had low levels of hepatitis C RNA. Of this group of 13 cryoglobulin-positive/hepatitis C antibody–negative patients, one had a mixed cryoglobulin, four had single-component cryoglobulins (IgM), and the others were trace positive. Four of five patients with single-component IgM cryoglobulinss had Mspikes greater than 3.5 g/dL by SPE. Eight of nine patients with mixed cryoglobulins also had RF positivity with titers between 1:640 and 1:327,680. Six of six of these patients tested for hepatitis C antibody and RNA were

positive for both, including patient Sie (23CG) from 19701,10,11 (Table 2). One patient (18CG) with a mixed cryoglobulin had a 4.1-g/dL IgM, kappa spike, but was RF-negative. This patient also was negative for hepatitis C (antibody and RNA). The relationship between mixed cryoglobulinemia, HCV, and Waldenstrom’s macroglobulinemia was well illustrated by a 52-yearold man (patient 21CG) with hepatitis C cirrhosis who was referred for liver transplant evaluation. He was found to have a 4.7-g/dL IgM, kappa serum M-spike, mixed cryoglobulinemia with a RF titer of 1:81,920, serum viscosity 6.3, hepatitis C antibody and RNA, and 20% marrow lymphoplasmacytosis. Of the 19 patients positive for RF, mixed cryoglobulins were present in eight. Three of the RFpositive patients also had ANA activity. Sera with trace or single-component cryoglobulins were RFnegative (Table 2). None had CA reactivity.

HUMAN MONOCLONAL MACROGLOBULIN ANTIBODIES

321

antigen ranged from 1:512 to 1:65,536. Reactivity against the “i” antigen was present in four patients but at much lower titer than against “I” antigen. None of the CA patients had ANA or DNA antibody activity. Reactivity to Neural Antigens

Fig 1. Sie serum (23CG) at 4°C and 37°C on 9/20/02. This specimen was collected on 3/30/70. The antibody activity of this Waldenstrom’s macroglobulin is shown in Table 1 and described in references 1, 10, and 11.

Cold Agglutinins Ten patients had positive monoclonal IgM CA (Table 3). All had evidence of immunohemolytic anemia. Nine of the 10 had kappa L chains and had M-spikes less than 1 g/dL. None were cryoglobulin- or RF-positive. CA titers against the “I”

All sera were tested for IgM binding to a series of glycosaminoglycan oligosaccharides, glycolipids, and glycoprotein antigens. Binding to one or more antigens was found in 75 (43%). Glycosaminoglycan oligosaccharides were the most common antigenic targets, with high titer IgM binding found in 56 sera (33%). IgM binding to antigens having known associations to polyneuropathies occurred in 20 patients (12%), seven of whom had neurologic symptoms or findings. IgM binding to GalNAc-GD1a ganglioside was the most common finding (11 patients; 6%). IgM binding to myelinassociated glycoprotein (MAG) was found in seven sera (4%). Other neuropathy-associated glycolipid targets included sulfatide (four patients; 2.3%), GM1 ganglioside (three patients; 1.7%), GM2 ganglioside (one patient; 0.6%), and GD1b ganglioside (one patient; 0.6%). IgM binding to tubulin was found in one patient (0.6%). IgM binding to another glycolipid that is not generally considered as a target of IgM M-proteins, GQ1b ganglioside, was found in three patients (1.7%).

Table 3. Cold Agglutinin Patients Patient

Year

Serum M-Spike

IgM L Chain

1CA 2CA

1985 1998

0.4 0.8

Kappa Kappa

3CA

1998

0.6

Kappa

4CA

1998

0.4

Kappa

5CA

2001

0.3

Kappa

6CA 7CA 8CA 9CA 10CA

1984 1985 1990 1977 1993

2.6 None None 3.5 None

Kappa Kappa Kappa Lambda Kappa

CA Titer

HgB

HCT

Retic

(⫹) I⫽1:65,536; i⫽1:512 I⫽1:4,096; i⫽1:32 I⫽1:8,192; i⫽1:1,024 I⫽1:16,384; i⫽1:256 1:8,192 1:512 1:32,768 1:1,034 1:8,192

12.0 9.1

34 21

— 10.2

—

19

4.5

11.3

33

—

12.6

36

6.2

9.7 13.0 9.8 11.5 8.7

28 38 29 33 22

5 10.4 3.8 0.6 17.0

Abbreviations: Hgb, hemoglobin; HCT, hematocrit (%); Retic, reticulocyte count (%).

322

STONE ET AL

Table 4. Antigen-Binding Activities of Some Waldenstrom Macroglobulins and Human Myeloma Proteins Foreign Antigens

Antigens of Autogenous Origin

Streptolysin O Staphylolysin Strep hyaluronidase Strep Dnase Phosphoryl choline ␣2-Macroglobulin (horse) Mycobacterial glycolipids Klebsiella polysaccharides Cytomegalovirus Drugs Sulfa Contrast medium Synthetic antigens Nitrophenyl derivatives

Human red blood cell antigens: polysaccharides IgG (RF) Serum albumin Transferrin Thyroglobulin Nucleoprotein-associated antigens ANA snRNP Sm SS-A/Ro SS-B/La Histones ds and ss DNA Membrane-associated antigens Phospholipid (cardiolipin) Lipoprotein-associated antigens ␣-, ␤-lipoproteins LDL, VLDL, HDL Blood coagulation-associated antigens Glycoprotein IIIa Fibrin monomer Factor VIII Neural antigens (myelin-associated glycoprotein and others) Cytoskeletal antigens Actin Myosin Tubulin Low-molecular-weight ligands Riboflavin Heparan sulfate Ions Acid polysaccharides (chondroitin sulfate, heparin)

Abbreviations: LDL, low-density lipoprotein; VLDL, very–low-density lipoprotein; HDL, high-density lipoprotein. Data from references 1, 24 –26.

Antinuclear Antibodies Eighteen patients had positive ANAs, with titers ranging from 1:40 to 1:10,240. As noted, three patients had both RF and ANA, and two of these three had biclonal gammopathies. Anti-DNA Activity None of the 167 samples tested exhibited antiDNA antibodies. DISCUSSION

The presence of antibody activity in serum from a patient with monoclonal gammopathy does not necessarily indicate that the M-protein possesses the activity. Criteria to establish that monoclonal

immunoglobulins are antibodies have been discussed.1,5-11 That some monoclonal macroglobulins and myeloma proteins satisfy these criteria has been documented for more than three decades. Most of the antigens identified have been of autologous or bacterial origin (Table 4).1,24-26 Within the macroglobulin group, CAs, mixed cryoglobulins, and reactivity to myelin-associated glycoprotein or other neural antigens have been most frequently identified. These three groups together accounted for 55% of the serum samples in this series. Other autoantigens were detected in 16% of specimens. It is possible that the other third of monoclonal IgMs are antibodies to antigens that have not been identified. The high in-

HUMAN MONOCLONAL MACROGLOBULIN ANTIBODIES

cidence of antibody activity in this study lends support to the view that the term “paraprotein” is a misnomer.6 Rather, monoclonal immunoglobulins appear to be “homogeneous representatives of a heterogeneous population.”27 It may be more difficult to detect antigens that react with monoclonal antibodies as compared with those reacting with usual polyclonal immune sera.6,9 Thus, negative assays that measure the consequences of primary antigen-antibody binding, such as agglutination or precipitation, do not rule out antibody activity. With respect to CAs, cryoglobulins, and neural antigens, the clinical manifestations of hemolytic anemia, cold sensitivity, and polyneuropathy, respectively, provide important clues to identification of the responsible antigens. Such patients often present at an earlier stage of their lymphoproliferative disease and have small serum IgM M-components even though they are symptomatic. Thus clinical manifestations due to antibody activity of monoclonal macroglobulins have important implications for diagnosis and therapy.28,29 In addition to signs and symptoms related to cold exposure, cryoglobulinemia has an adverse effect on serum viscosity.10,11 During the past decade, Agnello established the relationship between mixed cryoglobulins and HCV.30 The viral infection is associated with a B-cell lymphoproliferative response of limited or overtly malignant degree.28 In our study, six of six patients with mixed cryoglobulins and positive RF had evidence of HCV infection (antibody and RNA) (Table 2). These results support the importance of a relevant antibody response in the pathogenesis of Waldenstrom’s macroglobulinemia. Antigenic targets of serum IgM were found in a surprisingly high proportion of sera in this series of patients with monoclonal macroglobulinemia. Previously undescribed antigens included a series of glycosaminoglycan oligosaccharides. The specific antigenic epitopes in these oligosaccharides are currently being evaluated. For neuropathy-associated antigens, the most frequent target of IgM (6% of sera) was GalNAc-GD1a ganglioside. Monoclonal IgM binding to this ganglioside has been described in a sensory ataxic demyelinating polyneuropathy that improves after treatment with intravenous immunoglobulin, but is poorly responsive to other immunomodulating therapies. IgM binding to myelin-associated glycoprotein,

323

which is associated with a syndrome of sensory neuropathy, gait ataxia, tremor and distal leg weakness, was surprisingly uncommon (4% of sera). The frequency and type of neuropathy and systemic syndromes in patients with and without IgM binding to specific antigens will require examination and testing of these patients. Chemically, some monoclonal immunoglobulins appear to be polyreactive. A conventional (polyclonal) antibody response gains its high specificity from the unique mixture of antibody structures that form the population. Individual, ie, monoclonal, antibody molecules are less specific and, therefore, often appear to be polyreactive.1,9,31-34 Polyreactivity was observed in this study but its incidence was not established. From a biological view, the characterization of antigen-binding activity in monoclonal macroglobulins may provide further insight into the pathogenesis of monoclonal gammopathies. ACKNOWLEDGMENT We thank Drs Rana Saad and George Netto for performing the hepatitis C RNA determinations.

REFERENCES 1. Stone MJ: Monoclonal antibodies in the prehybridoma era: A brief historical perspective and personal reminiscence. Clin Lymphoma 2:148-154, 2001 2. Christenson W, Dacie J: Serum proteins in acquired hemolytic anemia (auto-antibody type). Br J Haematol 3:153164, 1957 3. Kritzman J, Kunkel H, McCarthy J, et al: Studies of a Waldenstro¨ m-type macroglobulin with rheumatoid factor properties. J Lab Clin Med 57:905-917, 1961 4. Waldenstro¨ m J, Winblad S, Hallen J, et al: The occurrence of serological “antibody” reagins or similar gamma-globulins in conditions with monoclonal hypergammaglobulinemia, such as myeloma, macroglobulinemia, etc. Acta Med Scand 176:619-631, 1964 5. Metzger H: Characterization of a human macroglobulin. V. A Waldenstro¨ m macroglobulin with antibody activity. Proc Natl Acad Sci USA 57:1490-1497, 1967 6. Stone M, Metzger H: The valence of a Waldenstro¨ m macroglobulin antibody and further thoughts on the significance of paraprotein antibodies. Cold Spring Harbor Symp Quant Biol 32:83-88, 1967 7. Stone MJ, Metzger H: Study of macromolecular interactions by equilibrium molecular sieving. J Biol Chem 243:50495055, 1968 8. Stone MJ, Metzger H: Binding properties of a Waldenstro¨ m macroglobulin antibody. J Biol Chem 243:5977-5984, 1968 9. Stone MJ, Metzger H: The specificity of a monoclonal macroglobulin (gamma M) antibody: Reactivity with primate gamma G immunoglobulins. J Immunol 102:222-228, 1969

324

10. Stone MJ: Studies on monoclonal antibodies. I. The specificity and binding properties of a Waldenstro¨ m macroglobulin with anti-␥G activity. J Lab Clin Med 81:393-409, 1973 11. Stone MJ, Fedak JE: Studies on monoclonal antibodies. II. Immune complex (IgM-IgG) cryoglobulinemia: The mechanism of cryoprecipitation. J Immunol 113:1377-1385, 1974 12. Attaelmannan M, Levinson SS: Understanding and identifying monoclonal gammopathies. Clin Chem 46:12301238, 2000 13. Ritzmann SE, Daniels JC: Serum protein electrophoresis and total serum proteins, in Ritzman SE (ed): Serum Protein Abnormalities: Diagnostic and Clinical Aspects (vol 1). Boston, MA, Little Brown, 1979, pp 3-25 14. Whicher JT, Price CP, Spencer K: Immunoephelometric and immunoturbidimetric assays for proteins. Crit Rev Clin Lab Sci 18:213-260, 1983 15. Brouet J, Clauvel JP, Danon F, et al: Biologic and clinical significance of cryoglobulins. Am J Med 57:775-778, 1974 16. Nakamura RM, Tucker ES: Immune complex diseases, in Ritzmann SE, Daniels JC (eds): Serum Protein Abnormalities: Diagnostic and Clinical Aspects (ed 1). Boston, MA, Little Brown, 1979, pp 295-330 17. Ulvestad E, Berentsen S, Bo K, et al: Clinical immunology of chronic cold agglutinin disease. Eur J Haematol 63:259266, 1999 18. Pestronk A, Choksi R, Yee WC, et al: Serum antibodies to heparan sulfate glycosaminoglycans in Guillain-Barre´ syndrome and other demyelinating polyneuropathies. J Neurolimmunol 91:204-209, 1998 19. Pestronk A, Li F, Bieser BS, et al: Anti-MAG antibodies: Major effects of antigen purity and antibody cross-reactivity on ELISA results and clinical correlation. Neurology 44:11311137, 1994 20. Chan EKL, Andrade LEC: Antinuclear antibodies in Sjogren’s syndrome. Rheum Dis Clin North Am 18:551-570, 1992 21. Chubick A, Sontheimer RD, Gilliam JN, et al: An appraisal of tests for native DNA antibodies in connective tissue diseases. Ann Intern Med 89:186-192, 1978 22. Ross RS, Viazov SO, Hoffmann S, et al: Performance characteristics of a transcription-mediated nucleic acid ampli-

STONE ET AL

fication assay for qualitative detection of hepatitis C virus RNA. J Clin Lab Anal 15:308-313, 2001 23. Trimoulet P, Halfon P, Pohier E, et al: Evaluation of the VERSANT HCV RNA 3.0 assay for quantification of hepatitis C virus RNA in serum. J Clin Microbiol 40:2031-2036, 2002 24. Merlini G, Farhangi M, Osserman EF: Monoclonal immunoglobulins with antibody activity in myeloma, macroglobulinemia and related plasma cell dyscrasias. Semin Oncol 13: 360-365, 1986 25. Potter M: Myeloma proteins and antibodies, in Moulin AM, Cambrosio A (eds): Singular Selves. Historical Issues and Contemporary Debates in Immunology. Paris, France, Elsevier, 2001, pp 23-43 26. Pestronk A: Autoantibodies and immune polyneuropathies, in Katirji B, Kaminiski HJ, Preston DC, et al (eds): Neuromuscular Disorders in Clinical Practice. Boston, MA, Butterworth Heinemann, 2002, pp 48-63 27. Metzger H: Molecular versatility of antibodies. Immunol Rev 185:186-205, 2002. 28. Stone MJ: Commentary: What is Waldenstro¨ m’s macroglobulinemia? Clin Lymphoma 1:44-45, 2000 29. Stone MJ: Myeloma and macroglobulinemia: What are the criteria for diagnosis? Clin Lymphoma 3:23-25, 2002 30. Agnello V: The etiology and pathophysiology of mixed cryoglobulinemia secondary to hepatitis C viral infection. Springer Semin Immunopathol 19:111-129, 1997 31. Dighiero G, Guilbert B, Fermand JP, et al: Thirty-six human monoclonal immunoglobulins with antibody activity against cytoskeleton proteins, thyroglobulin, and native DNA: Immunologic studies and clinical correlations. Blood 62:264270, 1983 32. Chen ZJ, Sheeler J, Notkins AL: Antigen-binding B cells and polyreactive antibodies. Eur J Immunol 25:579-586, 1995 33. Melero J, Aguilera I, Mageed RA, et al: The frequent expansion of a subpopulation of B cells that express RF-associated cross-reactive idiotypes: Evidence from analysis of a panel of autoreactive monoclonal antibodies. Scand J Immunol 48:152-158, 1998 34. Aguilera I, Melero J, Nunez-Roldan A, et al: Molecular structure of eight human autoreactive monoclonal antibodies. Immunology 102:273-280, 2001