IMMUNOPATHOLOGY
Original Article
Incidence of Antiphospholipid Antibodies in
Patients With Monoclonal Gammopathy
of Undetermined Significance
J. JAROSLAV STERN, MD, 1 RONALD H. NG, P H D , 2 DOUGLAS A. TRIPLETT, MD, 3
AND JOHN A. McINTYRE, P H D , M R C P A T H 1
The incidence of antiphospholipid antibodies in patients with monoclonal gammopathy of undetermined significance (MGUS) was studied.
Antiphospholipid antibodies were measured in the sera of 93 patients
(49 women, 44 men; mean age 701± 21] years) with MGUS by using an
enzyme-linked immunosorbent assay (ELISA). The phospholipids
tested were cardiolipin (CL), phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylglycerol (PG), phosphatidic acid (PA), phosphatidylcholine (PC), and phosphatidylethanolamine (PE).
Positive results were defined as a value higher than the number of
multiples of the mean per phospholipid, which included 76 of 80 (95%;
20 age-matched) control individuals.
The immunoglobulin-G (IgG) or IgM antiphospholipid antibodies
isotype varied among the patients, as did the phospholipid specificity.
For IgG, PI was found elevated in 32% of the MGUS samples, whereas
the other phospholipid antigens ranged from 9% to 15%. The percentage of patients with IgM antiphospholipid antibodies was higher. The
authors observed PS, PI, PA, and PC as positive in 45%, 35%, 25%, and
25% of patients, respectively. Of the 12 sera studied for IgA isotype,
three (25%) were positive for PS, six (50%) for CL, and none for PE.
Patients with MGUS manifested a significantly higher (P < .01)
incidence of antiphospholipid antibodies in their blood than did the
control persons. No difference in the incidence of antiphospholipid antibodies was seen between younger and older (age-matched) control patients. No correlation was found between serum levels of immunoglobulins and optical density reading of the blank plates used as ELISA
controls. (Key words: Cardiolipin; Cofactor, M-Protein; Paraprotein)
Am J Clin Pathol 1994;101:471-474.
The term monoclonal gammopathy of undetermined significance (MGUS), or benign monoclonal gammopathy, refers to
a condition characterized by the presence in serum of a monoclonal protein (M-protein) in persons without evidence of multiple myeloma, macroglobulinemia, amyloidosis, or other related diseases.1 The M-protein is an immunoglobulin (Ig) that
persists at an elevated and constant concentration over many
years. Monoclonal gammopathy of undetermined significance
becomes detectable in the serum when a single clone of mature
and differentiated B lymphocytes has proliferated to the degree
that gammaglobulin production is sufficient for the sensitivity
of immunochemical diagnostic procedures. 2 The ability of
monoclonal antibodies to bind to autoantigens, such as cytoskeletal proteins, thyroglobulins, native DNA, and insulin, is
unusual. 3,4 Up to 10% of MGUS sera will show autoantibody
activity, without autoimmune disease.5 The purpose of this
study was to determine the incidence of antiphospholipid antibodies (aPA) in a population with MGUS.
MATERIALS AND METHODS
Patients
and Control
Individuals
The study group consisted of 93 patients, 49 women and 44
men, with a laboratory diagnosis of MGUS. The patients were
aged 49-91 years (mean, 70 years). Sera were collected and
stored at —70 °C until used. Electrophoresis testing of the sera
found eight patients with biclonal and 85 with monoclonal
gammopathies. Eighty healthy persons, 60 aged 25-45 years
and 20 aged 50-85 years, were recruited as controls. No statistical difference was noted in the presence of aPA between these
two control groups.
Antiphospholipid
Antibody
Immunoassay
Enzyme-linked immunosorbent assays (ELISAs), modified
from that described by Triplett and colleagues,6 were established for antibodies to seven different phospholipids: cardiolipin (CL), phosphatidylserine (PS), phosphatidylethanolamine
(PE), phosphatidylinositol (PI), phosphatidic acid (PA), phosFrom the 'Center for Reproduction and Transplantation Immunol- phatidylcholine (PC), and phosphatidylglycerol (PG). All phosogy and ^Department of Pathology and Laboratory Medicine, Metho- pholipids used were obtained from Sigma Chemical Co. (St.
dist Hospital of Indiana, Indianapolis, Indiana; and 3Ball Memorial
Louis, MO). Thirty microliters of each phospholipid, diluted to
Hospital, Muncie, Indiana.
50 mg/mL in methanolxhloroform 3:1, were applied to the
wells of a 96-well flat-bottom polyvinyl plate and evaporated to
Manuscript received February 9, 1993; revision accepted June 3,
dryness under N 2 . Afterward, the wells were washed three
1993.
times with 200 nL TRIS-buffered saline (TBS) (Sigma). One
Address reprint requests to Dr. Mclntyre: Center for Reproduction
hundred microliters of 10% bovine serum albumin (BSA) were
and Transplantation Immunology, Methodist Hospital of Indiana,
then added as a block to each well and incubated for 1 hour.
1701 North Senate Boulevard, Indianapolis, IN 46202.
471
472
IMMUNOPATHOLOGY
Original Article
TABLE 1. OO READINGS USED AS THRESHOLD VALUES
TO DETERMINE POSITIVE APA RESULTS
IgM
PL
IgG
Antigen
95%
MoM
95%
MoM
95%
MoM
PS
CL
PG
PI
PE
PC
PA
0.156
0.315
0.244
0.174
0.140
0.208
0.160
3
3
2
3
4
4
4
0.066
0.072
0.186
0.144
0.132
0.174
0.096
3
3
3
3
3
3
4
0.054
0.066
3
3
TABLE 2 THE PERCENTAGE OF APA POSITIVE SERA
IN MGUS PATIENTS
IgA
Antigen
Antibody*
PS
CL
PG
PI
PE
PC
PA
IgG
9%
4-6
45%
4-9
25%
4-8
9%
4-16
5%
4-19
50%
3-10
12%
3-13
10%
4-8
NT
NT
32%
4-20
35%
4-8
NT
NT
13%
5-49
10%
4-8
0
9%
5-9
25%
4-11
NT
NT
15%
5-11
25%
5-8
NT
NT
t
IgM
0.140
4
t
IgA
t
Multiples of the mean (MoM) that are shown included 95% of the individuals in the control
groups.
The wells were washed three times with TBS, and 50 iiL of the
test and control sera diluted to 1:100 in TBS containing 10%
adult bovine serum (ABS) were added in triplicate and incubated for 1 hour at room temperature. The wells were then
washed three times in TBS. Fifty microliters of the appropriate
alkaline phosphatase conjugated antihuman immunoglobulin
(IgG and IgM, Sigma) in 1:1000 dilution of 10% ABS/TBS
were added to each well and incubated for 1 hour at room
temperature. After three washes, the color reaction was developed for 1 hour with 50 ML paranitrophenyl phosphate at 37 °C
in darkness. The reaction was stopped by the addition 75 fiL of
3 mol sodium hydrochloride. Color development was measured by reading the optical density through a 405-nm filter
with a Biomek 1000 Workstation (Beckman, Arlington
Heights, IL). Four controls were run with each plate: 1, negative; 2, weak positive; 3, medium positive; and 4, strong positive. The IgG control sera were selected based on validation of
our aPA ELISA by using a set of "standards" available from
Dr. Nigel Harris, University of Louisville, Kentucky. Weak,
medium, and strong positive controls were obtained by threepoint dilutions of an aPA positive sample. The control sera
were stored in aliquots at - 70 °C, and a freshly thawed aliquot
was used for each assay. A blank plate (no antigen) for each
isotype was done for each sample tested. Twelve patient samples with high concentrations of IgA were studied for the presence of IgA, aPA to CL, PS, and PE and compared with a
normal population of 255 controls.
NT = not tested.
* Statistical differences from controls: IgG, P = 0.01; IgM, P " 0.0002; IgA, P = <0.001.
f Ranges of the multiples of the mean (MoM) above control group threshold.
OVA). Differences with P values < .05 were considered statistically significant.
RESULTS
Our ELISA established the negative/positive threshold
value, with a 95% confidence interval for IgG, IgM, and IgA
aPA isotypes. Table 1 shows the MoM calculated from 80 control persons that were determined for each phospholipid antigen for IgG and IgM and the MoM for IgA calculated for our
previously established normal control population of 255 persons.
Table 2 shows the percentage of aPA positive sera from patients with MGUS. The IgG or IgM aPA isotype varied among
the patients, as did the phospholipid specificity. For IgG, PI
was elevated in 32% of the MGUS samples, whereas the other
phospholipid antigens ranged from 9% to 15%. The percentage
of patients with IgM aPA was higher. We observed PS, PI, PA,
and PC as positive in 45%, 35%, 257o, and 25% of patients,
respectively. Of the 12 sera studied for IgA isotype, three (25%)
were positive for PS, six (50%) for CL, and none for PE. Table 2
also provides the range of the MoM calculated for the MGUS
patients.
mg/dl
*
2800 -:
2400 H
Monoclonal proteins in serum samples were identified by
immunofixation electrophoresis with the Paragon System
(Beckman Instruments, Inc., Brea, CA).7 The concentration of
each immunoglobulin and light chain was quantified with
monospecific antisera and standards using the Behring Nephelometer (Behring Diagnostics Inc., Somerville, NJ).8
*
2000 -•
+
•
»
+
1200 -
*
800 400 j
0 J SL.
IjG
+
1600 \
• 't
- ** •
**. * -
•
•
•
IgA
II
.•
^fO.
•
I,V:I
Jt0.05
• t
, ^C * i
0.1
*
1
0.15
IgM
\ . '.*
0.2
0.25
O.D
IMMUNOGLOBULIN
+lgG XlgM QlgA
Statistical Analysis
Patients results are reported as multiples of the mean (MoM)
of 80 healthy persons for IgG and IgM and 255 normal controls
for IgA tested on each combination of phospholipid antigen. A
positive value is the MoM that is greater than or equal to 95%
of the healthy persons tested. Data are expressed as percentages
of positive patients, then analyzed by analysis of variance (AN-
•
-
3200 -
Electrophoresis and Nephelometry
—
FIG. 1. Correlation of the MGUS serum sample between the nephelometry quantitation of the IgG isotype and the corresponding optical density reading of the ELISA background (no antigen) control plate. The
horizontal lines indicate the upper limits for normal immunoglobulin
values.
A.J.C.P. • April 1994
STERN ET AL.
Antiphospholipid Antibodies in MGUS
mg/dl
473
Twenty-two sera (24%) studied were positive to two or more
phospholipids, and six sera were positive to more than four
2800 -i
phospholipids. Two of the 93 sera were positive with two differ\
2400 -j
ent isotypes to 9 and 11 phospholipid antigens, respectively,
„
2000 -i
and these sera were from biclonal gammopathy patients. The
*
1600 H
phenomenon of multispecific antibodies is now well recog1200 -j *
nized and may have important clinical implications in autoim•
\
800 -j
munity. It may be a combining site region that has different
adjacent
or overlapping subregions capable of recognizing dif400 -j <"S
* *w
ferent epitopes or various unrelated proteins that exhibit a simio3
0.05
0.1
0.15
0.2
0.25
lar surface epitope determinant."
Some investigators have observed an increased incidence of
O.D
aPA in elderly populations.16,17 As reported in other studies,18"23
IMMUNOGLOBULIN
+ lgG XlgM olgA
however, we did not detect a difference in the presence of aPA
when we compared the results of the younger (60 persons ages
FIG. 2. Correlation of the MGUS serum sample between the nephelom- 25-45 years) and older (20 persons aged 50-85 years) control
etry quantitation of the IgM isotype and the corresponding optical
populations.
density reading of the ELISA background plate (no antigen) control.
Our report differs from another study,24 wherein a direct
The horizontal lines indicate the upper limits for normal immunoglobcorrelation
between high concentrations of proteins in sera was
ulin values.
described as an explanation for the high background (blank
plate) values in the ELISA. As shown in Figures 1 and 2, we did
not find a correlation between the immunoglobulin concentrations as quantified by nephelometry and the optical density
The increased immunoglobulin concentrations characterisreading of nonantigen-containing (blank) plates in the ELISA.
tic of MGUS sera provided the opportunity to evaluate what, if
Furthermore, we did not observe any correlation between the
any, correlation existed between serum immunoglobulin levels
ELISA background values and the antiphospholipid positivity.
and the blank plate controls, (ie, nonphospholipid antigenThis raises new questions regarding the mechanism(s) of reaccoated wells). Figures 1 and 2, illustrate how the serum IgG and
tivity in blank plates and the interpretations offered. One possiIgM concentrations of immunoglobulins, as measured by the
bility, for which we have preliminary data, is that the high
nephelometer, correlated with the optical density reading of
background IgM binding is caused by a property shared by
the blank plate controls used in the ELISA.
most human IgM antibodies processing VHIII H chain variable
No correlation was found between immunoglobulin concenregions, which are absent from IgM bearing the VHHI and VHII
tration and blank plate reactivity. Similarly, no correlation was
subgroup markers.25 The IgM expressing VHIII H chain binds
observed between blank plate reactivity and antiphospholipid
specifically to protein A. We have found that adsorptions of
positivity.
patient sera manifesting high IgM background reactivity with
protein A abolishes their background reactivity (Mclntyre and
Wagenknecht, unpublished). If these initial observations hold
DISCUSSION
true for other patients with high background values, this particMonoclonal gammopathy of undetermined significance is
ular IgM VH gene sequence may have been conserved within
the most common type of M-protein detected by routine serum
the species and may have some as yet undefined important
protein electrophoresis. It occasionally has been observed as
survival value that is not related to antigenic specificity, per se,
part of an infectious process, in a postimmunization state, and
but to some other immunoregulatory function. Additional insubsequent to drug therapy. These M-proteins initially were
vestigation is needed in this area.
believed to be different structurally from normal immunoglobIn our study group, some of these M-proteins may have speciulins and thought to have no specific antigenic target. We now
ficity to phospholipids; however, we have no data to support
know, however, that a large number of human monoclonal
immunoglobulins have specificity to particular proteins, such
this. In Table 2, the patient who has an IgG anti-PE of 49 MoM
as insulin, albumin, and transferrin.4,5,9"12
would be a likely candidate to screen for aPA monoclonality,26
although it is not unusual for aPA titers in non-MGUS patients
A higher incidence of aPA has been found in the MGUS
to exceed 40 MoM. Nonetheless, antisera of such high titer are
serum samples compared with the normal control population,
useful for dissecting the role of cofactor requirements in the
of which 20 were age-matched. To our knowledge, this is the
aPA ELISA. For example, this particular MGUS patient's antifirst study to demonstrate the aPA incidence to seven different
PE reactivity is dependent on the presence of ABS as a source
phospholipid antigens in a MGUS population. We found that a
of cofactor in the ELISA. No reactivity is found when the ABS
given patient can be positive concomitantly to several phosphois replaced with BSA. Moreover, this PE cofactor is not ^-glylipids and with more than one isotype. Our MGUS population
consisted of patients with diverse diagnoses, eg, cardio- and
coprotein I, the molecule that serves as cofactor for anionic
cerebrovascular diseases, fractures, and infections. In the literaphospholipids in the ELISA.27 Identification of the PE cofactor
ture, several references demonstrate an association between
is in progress. Indeed, many plasma proteins probably serve a
monoclonal antibodies and an autoantibody activity; of these,
phospholipid cofactor function. Binding of these cofactors to
12 14
however, few have clinical manifestations. " Our data supdifferent phospholipids results in antigenic changes. This pheport previous reports of such autoantibody reactivities, but we
nomenon occurs for the /32 glycoprotein I molecule,28 and such
do not know if these patients had MGUS before presentation of
changes may be responsible for the stimulation and production
their primary diseases or secondary to their primary diseases.
of autoimmune aPA.
3200 1
IjG
•
•
•
„
•
, — , — .
IgA
t
IflM
;
I
I
.
I
1
1
1
Vol. 101 - N o . 4
474
IMMUNOPATHOLOGY
Original Article
CONCLUSIONS
Patients with MGUS manifest a significantly higher incidence of aPA in their blood than do age-matched control individuals. We did not find a difference in the incidence of aPA
between older and younger control individuals. We observed
that the background (blank plate) optical density readings in
the ELISA were not related to the concentrations of immunoglobulins in the sera of these MGUS patients, and they were
not correlated with the antiphospholipid positivity.
14.
REFERENCES
17.
1. Kyle RA, Lust JA. Monoclonal gammopathies of undetermined
significance. Semin Hematol 1989;26:176-200.
2. Radl J. Benign monoclonal gammopathy (BMG). Curr Top Microbiol Immunol 1986; 132:221-224.
3. Dighiero G, Guilbert B, Fermand JP, Lymberi P, Danon f, Avrameas S. Thirty-six human monoclonal immunoglobulins with
antibody activity against cytoskeleton proteins, thyroglobulin,
and native DNA: Immunologic studies and clinical correlations. Blood 1983;62:264-270.
4. Wasada T, Eguchi Y, Takayama S, Yao K, Hirata Y, Ishii S. Insulin autoimmune syndrome associated with benign monoclonal
gammopathy. Diabetes Care 1989;12:147-150.
5. Guilbert B, Dighiero G, Avrameas S. Naturally occurring antibodies against nine common antigens in human sera. J Immunol
1982;128:2779-2787.
6. Triplett DA, Brandt JT, Musgrave KA, Orr, CA. The relationship
between lupus anticoagulants and antibodies to phospholipid.
JAMA 1988;259:550-554.
7. Sun T, Lien YY, Degran T. Study of gammopathies with immunofixation electrophoresis. Am J Clin Pathol 1979;72:5-11.
8. Salden HJM, Bas BM, Hermans TH, Janson PCW. Analytical performance of three commercially available nephelometers compared for quantifying proteins in serum and cerebrospinal fluid.
Clin Chem 1988;34:1594-1596.
9. Osterland CK. Monoclonal gammopathies—their identification
and biological significance - critical review. Clin Chim Acta
1989;180:1-22.
10. Buskila D, Abu-shakra A, Amital-Teplizki A, et al. Serum monoclonal antibodies derived from patients with multiple myeloma
react with mycobacterial phosphoinositides and nuclear antigens. Clin Exp Immunol 1989;76:378-383.
11. Merlini G, Farhangi M, Osserman EF. Monoclonal immunoglobulins with antibody activity in myeloma, macroglobulinemia and
related plasma cell dyscrasias. Semin Oncol 1986; 13:350-365.
12. Buskila D, Weigel D, Shoenfeld Y. The detection of anti-Ro/55-A
and anti-La/55-B activity of human serum monoclonal immunoglobulins (monoclonal gammopathies). Hum Antibody Hybridomas 1992;3:75-80.
13. Krause T, Cohen J, Blank M, et al. Distribution of two common
idiotypes of anticardiolipin antibodies in sera of patients with
15.
16.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
primary antiphospholipid syndrome, systemic lupus erythematosus and monoclonal gammopathies. Lupus 1992; 1:91-96.
Potter M. Myeloma proteins (M-components) with antibody like
activity. N Engl J Med 1971;284:831-834.
Richards F, Konigsberg W, Rosenstein W, Varga J. On the specificity of antibodies. Biochemical and biophysical evidence indicates the existence of polyfunctional antibody combining regions. Science 1975; 187:130-136.
Fields RA, Toubbeh H, Seoules RP, Bankhurst AD. The prevalence of anticardiolipin antibodies in a healthy elderly population and its association with antinuclear antibodies. J Rheumatol 1989; 16:623-625.
Manoussakis MN, Tzioufas AG, Silis MP, Parge PJE, Goudevenos
J, Moutsopoulas HM. High prevalence of anticardiolipin and
other autoantibodies in a healthy elderly population. Clin Exp
Immunol 1987;69:557-565.
Gordon J, Rosenthal M. Failure to detect age-related increase of
non-pathological autoantibodies. Lancet 1984;i:231.
Goodwin JS, Searles RP, Tung KSK. Immunological responses of
a healthy elderly population. Clin Exp Immunol 1982;48:403410.
Klemp P, Cooper RC, Strauss FJ, Jordaan ER, Przybojewski JZ,
Nel N. Anti-cardiolipin antibodies in ischemic heart disease.
Clin Exp Immunol 1988;74:254-257.
PandeyJP, Fudenberg HH, Ainsworth SK, LoadholtCB. Autoantibodies in healthy subjects of different age groups. Mech Ageing
Dev 1979;10:399-404.
DeCaterlina R, d'Ascanio A, Mazzone A, et al. Prevalence of anticardiolipin antibodies in coronary artery disease. Am J Cardiol
1990;65:922-923.
Kushner MJ. Prospective study of anticardiolipin antibodies in
stroke. Stroke 1990;21:295-298.
Cowchock S, Fort J, Munoz S, Norberg R, Maddrey W. False
positive ELISA tests for anticardiolipin antibodies in sera from
patients with repeated abortion, rheumatologic disorders and
primary biliary cirrhosis: correlation with elevated polyclonal
IgM and implications for patients with repeated abortion. Clin
Exp Immunol 1988;73:289-294.
Susso EH, Silverman GJ, Mannik M. Human IgM molecules that
bind staphylococcal protein A contain VHIII H chains. J Immunol 1989; 142:2778-2883.
Kabat EA. Overview: The quest for antibody homogeneity: The
sine qua non for structural and genetic insights into antibody
complementarity. In: Weir DM, ed. Handbook ofExperimental
Immunology. Boston: Blackwell Scientific, 1986, pp. 11.111.17.
Stern JJ, Roussev RG, Vanderpuye OA, Mclntyre JA. Non-/?2
Glycoprotein I "cofactor" dependence of human anti-phosphatidylethanolamine antibodies in the ELISA. J Clin Exp Rheumatol 1993; 10:644.
Wagenknecht DR, Mclntyre JA. Changes in #2 Glycoprotein antigenicity induced by phospholipid binding. Thromb Hemoslas
1993;69:361-365.
A.J.C.P. •April 1994