Application of the Immunoperoxidase Technic to Bone Marrow
Trephine Biopsies in the Classification of Patients
with Monoclonal Gammopathies
LOANN C. PETERSON, M.D., BRADLEY A. BROWN, M.D.,
JOHN T CROSSON, M.D., AND JEANETTE MLADENOVIC, M.D.
This study evaluated the utility of the immunoperoxidase method
as applied to bone marrow sections in the diagnosis of patients
with monoclonal gammopathies. Intracellular immunoglobulin
light chains were identified in fixed, decalcified bone marrow
biopsy sections from 66 patients with monoclonal proteins, using
an avidin-biotin-peroxidase complex immunoperoxidase method.
In all cases the predominant light chain identified in the bone
marrow biopsy correlated with the monoclonal light chain identified in the serum. In addition, a light chain ratio was defined
that correlated with the clinical diagnoses. The light chain ratios
were highest in patients with multiple myeloma and were significantly different from those with monoclonal gammopathy of
undetermined significance. There was no correlation between
level of serum monoclonal protein and light chain ratios. The
ratios were also high in patients with macroglobulinemia, primary
amyloidosis, and renal disease secondary to monoclonal proteins
but without overt myeloma. Determination of light chain ratios
differentiated patients with multiple myeloma from those with
monoclonal gammopathy of undetermined significance and helped
identify patients with end organ damage secondary to monoclonal
proteins but without overt myeloma. (Key words: Immunoperoxidase; Monoclonal gammopathy; Multiple myeloma; Monoclonal gammopathy of undetermined significance) Am J Clin
Pathol 1986; 85: 688-693
MONOCLONAL IMMUNOGLOBULINS have been
reported in 1% of the population older than age 25' and
in 3% older than age 70.' ,4 Most patients with monoclonal
proteins have a "benign monoclonal gammopathy" with
no evidence for a malignant disease of B-lymphocytes or
plasma cells. However, some patients originally classified
as having "benign monoclonal gammopathy" eventually
develop multiple myeloma, macroglobulinemia, or lymphoma. 8 This has led Kyle to suggest that the term
"monoclonal gammopathy of undetermined significance"
be used at the time of initial manifestation to better describe these patients.8
In most instances, patients with monoclonal gammopathy of undetermined significance can be distinguished
easily from those with multiple myeloma or other mono-
Received July 23, 1985; received revised manuscript and accepted for
publication September 20, 1985.
Address reprint requests to Dr. Peterson: Department of Pathology,
Hennepin County Medical Center, 701 Park Avenue, South, Minneapolis,
Minnesota 55415.
Departments of Pathology and Medicine, Hennepin County
Medical Center and Veterans Administration Medical Center,
Minneapolis, Minnesota
clonal diseases. There are, however, some patients who
are difficult to classify even after clinical, roentgenologic,
and bone marrow biopsy data are correlated. The purpose
of this study was to assess the utility of the immunoperoxidase method on bone marrow biopsy sections in the
classification of patients with monoclonal gammopathies.
Materials and Methods
Patient Population
The study population consisted of 73 patients who had
serum monoclonal immunoglobulins identified at Hennepin County Medical Center from January 1978 to July
1984. All patients had bone marrow aspirates and posterior iliac spine core biopsies obtained with the Jamshidi
needle7 after detection of the monoclonal protein.
Laboratory Methods
The monoclonal immunoglobulins were detected in the
serum, using standard immunoelectrophoretic technics.
The electrophoresis was performed on immunoelectrofilms (Kallestad, Austin, TX) at pH 8.6 in barbital buffer.
Sera were reacted with anti-total immunoglobulin, antiIgG, anti-IgA, anti-IgM, anti-kappa, and anti-lambda
(Kallestad) and the precipitin lines compared with those
produced by normal human serum control. When appropriate, the sera were reacted with anti-IgD, anti-IgE,
anti-free kappa, and anti-free lambda (Kallestad, Austin,
TX). If an IgM monoclonal protein was suspected, the
serum was reduced with 2-mercaptoethanol. The electrophoretic films were interpreted at the end of diffusion and
after staining with Coomassie Brilliant Blue Stain. Immunoglobulins were quantitated by nephelometry (Beckman Instruments, Brea, CA), using standard procedures.
In selected situations, the monoclonal protein or proteins were confirmed by immunofixation. Appropriately
688
vol. 85-No. 6
IMMUNOPEROXIDASE TECHNIC
689
Table 1. Distribution of Immunoglobulin Classes According to Disease Category (66 patients)
Number of Cells
Counted 1
Disease Category
Monoclonal gammopathy of
undetermined
significance
Multiple myeloma
Amyloidosis
IgM monoclonal
gammopathy
Extramedullary
plasmacytoma
Renal disease
Fanconi syndrome
Light chain nephropathy
Renal tubular acidosis
Lymphoma
Minimum
No.*
Mean
No.f
IgG
IgA
12
52
84
84
>100
MOO
16
5
1
5
6
2
20
86
66
84
66
>100
IgM
Biclonal
Triclonal
1
8
Kappa
Only
1
9
Lambda
Only
Total
2
2
2
24
23
5
1
9
1
1
1
1
1
1
1
1
1
7
66
1
>100
>100
Total
23
14
* Minimum number of positively reacting cells counted in an individual case.
diluted sera were electrophoresed on Panagel® (Worthington Diagnostic Systems, Freehold, NJ), then overlaid
for one to two hours with Sepraphore III® strips soaked
in the appropriate monospecific anti-serum. The Panagel
was then washed, dried, stained with Coomassie Blue,
and interpreted.
The bone marrow aspirates were processed according
to a previously published method.2 The trephine biopsies
were fixed in B-5 fixative for two hours, decalcified for 90
minutes in RDO (DuPage Kinetic Laboratories, Naperville, IL), embedded in paraffin in the conventional manner, and cut into 3-4-/*m sections. Intracytoplasmic kappa
and lambda light chains within plasma cells were detected
using an avidin-biotin-peroxidase complex immunoperoxidase method.6 Briefly, the sections were deparaffinized
and endogenous peroxidase was blocked by incubation
with a 0.3% aqueous H 2 0 2 solution for 10 minutes. The
sections were rinsed with phosphate-buffered saline (PBS),
flooded with diluted normal goat serum (Vector Laboratories, Burlingame, CA) for 20 minutes, incubated with
rabbit anti-human antibodies to either kappa or lambda
light chains (Dakopatts A/S, Copenhagen, Denmark; U.S.
agent, Accurate Chemical, and Scientific Corporation,
Westbury, NY) for 30 minutes, and rinsed with PBS. The
sections then were incubated with biotinylated secondary
goat anti-rabbit immunoglobulin (Vector Laboratories,
Burlingame, CA) for 15 minutes, rinsed with PBS, and
then incubated with the avidin-biotin-peroxidase complex (ABC) (Vector Laboratories, Burlingame, CA) for 15
minutes. The slides were rinsed with PBS, developed with
a diaminobenzidine (Polysciences, Warrington, PA) reaction, washed, dehydrated, and counterstained with he-
8
1
1
12
t Mean number of positively reacting cells counted per case.
matoxylin. Controls included established positive and
negative tissue sections, omission of one or more stages
in the procedure, and the use of nonimmune rabbit serum
in place of the primary antibody.
Analyses
The sections were examined without knowledge of the
clinical diagnosis. Plasma cells that reacted positively for
either kappa or lambda light chains were counted in ten
1,000X oil immersion fields (using 100X oil lens and 10X
eyepiece) from similar areas in the biopsy core sections.
The minimum number of positively reacting cells and the
mean number of positively reacting cells counted in each
disease category is shown in Table 1. A light chain ratio
for each patient was calculated as follows:
Light chain ratio
number of positively reacting plasma cells
for the predominant light chain
number of positively reacting plasma cells
for the minority light chain
Light chain ratios determined on sections from 13 patients
without monoclonal proteins were used as normal controls.
The patients' medical records were reviewed and the
patients were classified by clinical diagnoses without
knowledge of the immunoperoxidase results. Patients were
diagnosed as having multiple myeloma according to criteria reported by Greipp and Kyle.3 Patients with multiple
myeloma had more than 10% plasma cells in the bone
marrow and either lytic lesions or osteoporosis. Patients
690
PETERSON ET AL.
A.J.C.P. • June 1986
on each patient. When the data were analyzed, there was
no difference in results between the two ratios. Therefore,
all data are presented as light chain ratios.
Results
Of the initial 73 patients who had serum monoclonal
proteins and had undergone bone marrow biopsies, 66
had sufficient material to be included in the study. Four
patients were excluded because of insufficient core biopsy
material remaining for study; three were excluded because
of inadequate results with the immunohistochemical procedure. The distribution of the immunoglobulin classes
in the various disease categories is represented in Table
1. An example of the staining reaction for immunoglobulin in bone marrow sections by the immunoperoxidase
method is shown in Figure 1. The light chain of the
monoclonal protein identified in the serum in all patients
was the same as the predominant light chain identified in
the bone marrow core biopsy by the immunoperoxidase
technic.
FlG. 1. Positive reaction for kappa light chains by the immunoperoxidase method in a decalcified bone marrow biopsy from a patient with
multiple myeloma. Inset: No reaction in same biopsy for lambda light
chains (X400).
without bone lesions had more than 30% plasma cells in
the bone marrow; those with fewer than 10% plasma cells
had lytic bone lesions. Patients with smoldering multiple
myeloma9 had more than 10% plasma cells in the bone
marrow but no bone lesions or other evidence of multiple
myeloma. Patients with monoclonal gammopathy of undetermined significance had a monoclonal protein other
than IgM, fewer than 10% plasma cells in the bone marrow, and no evidence of a malignant plasmacellular or
lymphocytic process. The mean follow-up of these patients
was 25 months. The remaining patients were classified as
having other monoclonal gammopathies such as primary
amyloidosis, IgM monoclonal gammopathy, renal disease
associated with monoclonal proteins, extramedullary
plasmacytoma, or malignant lymphoma.
Logarithms of the light chain ratios were taken for statistical analysis; Mests were used to test mean differences
of values between groups. In addition to the light chain
ratio in which the predominant light chain always represented the numerator, a kappa/lambda ratio, in which
kappa always represented the numerator, was determined
Figure 2 summarizes the light chain ratios of the major
study groups. The light chain ratio of the 13 control patients without monoclonal proteins was less than or equal
to four, with kappa being the predominant light chain in
all cases. Therefore, a ratio of less than or equal to four
served as the normal control value.
Twenty-two of the 23 patients with multiple myeloma
had abnormal light chain ratios. All 22 patients had light
chain ratios of 16 or greater, defining a myeloma range.
Fifteen of the 22 patients had ratios that were 100 or
greater. The one myeloma patient with a normal ratio
had a triclonal gammopathy (IgG/c, IgGX and another
IgGX) involving both kappa and lambda light chains.
The light chain ratios of patients with monoclonal
gammopathy of undetermined significance were significantly different from those with multiple myeloma (P
< 0.001). Eight of the 24 patients with monoclonal gammopathy of undetermined significance had light chain ratios in the normal range (<4). Fourteen of the 24 patients
had light chain ratios that were above the range of the
control group but less than 16. One of these patients (ratio
8) developed multiple myeloma in six months, with the
light chain ratio of the bone marrow at the time of diagnosis of overt myeloma being 22. Two of the 24 patients
with monoclonal gammopathy of undetermined significance had ratios in the range seen in multiple myeloma
(greater than 16). One patient died of complications secondary to malignant hypertension three months after detection of the monoclonal protein; no autopsy was performed. The remaining patient was lost to follow-up.
Nine patients had IgM monoclonal proteins. Two of
these patients had Waldenstrom's macroglobulinemia;
both had light chain ratios greater than 16. Seven patients
had no bone marrow neoplasia. Five of the seven had
Vol. 85 • No. 6
IMMUNOPEROXIDASE TECHNIC
691
FIG. 2. Light chain ratios in control group and various disease categories. Darkly shaded area represents
range of control group; unshaded
area represents multiple myeloma
range; lightly shaded area represents
abnormal ratios below the myeloma
range.
Control
light chain ratios in the range of the control group (<4).
One had cold agglutinin disease of the elderly and a ratio
of 15. The remaining patient had biclonal proteins, IgM/c
and IgG/c, cryoglobulinemia, and a light chain ratio of
greater than 100.
There were five patients with primary amyloidosis; all
had lambda as the predominant light chain. Four of the
five had abnormal lambda to kappa light chain ratios,
with three being in the multiple myeloma range (16 or
greater). Only one patient had a normal ratio.
Three patients (data not shown in figure) with renal
disease and monoclonal proteins were studied. One patient had Fanconi's syndrome associated with kappa light
chains in the urine and 3% plasma cells in the bone marrow, with morphology that has been described in this syndrome. I0 One had renal tubular acidosis with monoclonal
kappa light chains in the serum and urine; this patient
had 12% plasma cells with no bone lesions and probably
represented multiple myeloma. The third patient had light
chain nephropathy with 5% plasma cells and no bone
lesions or other evidence of multiple myeloma. These
three patients had markedly abnormal ratios of greater
than 100, 60, and 50, respectively.
Two additional patients were studied. One had a lymphocytic lymphoma involving the bone marrow. The light
chain ratio of this patient was greater than 100, with the
positively staining cells including plasma cells and plasmacytoid lymphocytes within the lymphoma population.
The other patient had an extramedullary plasmacytoma;
the bone marrow was uninvolved morphologically with
multiple myeloma, and the light chain ratio was 10.
When the patients with multiple myeloma were studied
in greater detail, no significant correlation between the
Multiple
Myeloma
Monoclonal Gammopathy
of Undetermined
Significance
Primary
Amyloidosis
percentages of plasma cells in the bone marrow and the
light chain ratios (r = —0.03) was found. The light chain
ratios of patients with multiple myeloma and plasma cell
percentages of less than 10% were not significantly different from the light chain ratios of patients with plasma cell
percentages greater than 10% (Fig. 3). The one patient
with smoldering myeloma had a ratio of greater than 100.
In addition, one patient with nonsecretory myeloma, 40%
plasma cells in the bone marrow, hypogammaglobulinemia, and lytic bone lesions had a ratio greater than 100.
There was no significant correlation between level of
serum monoclonal protein and the light chain ratio, nor
was there a correlation between the level of IgGx and the
ratio, IgGX and the ratio, IgG/c plus IgGX and the ratio,
or IgA/c plus IgAA and the ratio.
Discussion
Many patients undergo bone marrow biopsies in an
attempt to determine the significance of a monoclonal
immunoglobulin identified in the serum or urine. An early
study, using the immunoperoxidase method to identify
intracellular immunoglobulin in paraffin-embedded bone
marrow aspirates from patients with monoclonal gammopathies, reported a high incidence of anomalous staining and, therefore, lack of clear correlation between the
monoclonal protein in the serum or urine and the immunoperoxidase staining pattern in the bone marrow.13
Pinkus and Said, however, reported excellent correlation
between the immunoglobulin detected in iliac crest biopsy
sections and the monoclonal protein identified in serum
or urine of patients with multiple myeloma and macroglobulinemia.12 Hitzmann and colleagues confirmed the
692
PETERSON ET AL.
A.J.C.P. .June 1986
100
0
~~
80 —
0
a>
s
>
* o
a> n
Q.2
TK
60
—
o
»
5 oc
«(Q
§
—
40 —
0
E c
0
FIG. 3. Light chain ratio versus plasma cell
percentage in bone marrow. There is no significant correlation between the two values. The
shaded area represents cases with fewer than
10% plasma cells in the bone marrow.
M
o
a.
20 —
o
0
0
s
§
8
0
o
I
20
0
I
I
I
40
60
80
•.
'
0
i
>100
Light Chain Ratio
specificity of the immunoperoxidase method and suggested its usefulness in evaluating patients with morphologically unusual myeloma and in the differential diagnosis of bone marrow plasmacytosis.5
The present study evaluated the utility of the immunoperoxidase method as applied to decalcified bone marrow core sections in the classification and diagnosis of
patients with monoclonal gammopathies. In all 66 cases
the predominant light chain identified in the bone marrow
biopsy correlated with the monoclonal light chain identified in the serum. In addition, a light chain ratio was
defined that correlated with the clinical diagnoses.
The light chain ratios were highest in patients with
multiple myeloma and were significantly different from
those with monoclonal gammopathy of undetermined
significance. Although two-thirds of the ratios in patients
with monoclonal gammopathy of undetermined significance were abnormal, very little overlap between these
two groups was seen. The ratio that appeared to best separate patients with multiple myeloma from those with
monoclonal gammopathy of undetermined significance
was 16. Within these two groups of patients, 92% with
ratios of 16 or greater had multiple myeloma and 96%
with ratios below 16 had monoclonal gammopathy of undetermined significance. Although not directly stated, data
reported by an Italian group 1 ' appear to support a similar
conclusion.
The high ratios seen in multiple myeloma were present
even in biopsies from patients with low percentages (less
than 10%) of plasma cells and in a single patient with
smoldering myeloma. This finding suggests that this technic may be useful when the diagnosis of myeloma is in
question. The ratio was also high (> 100) in a patient with
nonsecretory myeloma, confirming other reports that detection of immunoglobulin by this technic is useful when
no monoclonal protein is detected in serum or urine. 5 "
The significance of high or low ratios in patients with
monoclonal gammopathy of undetermined significance
is unclear. One patient with monoclonal gammopathy of
undetermined significance had an abnormal ratio and had
multiple myeloma develop within six months; further
studies, however, are needed to determine the significance
of an abnormal ratio in this group of patients and whether
a high ratio may be a predictor of progression to a malignant process.
The utility of the immunoperoxidase technic may be
extended to patients with other monoclonal diseases. For
example, two patients with Waldenstrom's macroglobulinemia had high ratios (above 16), while six of the seven
patients with no clinical malignancy had normal ratios.
In addition, four of the five patients with primary amyloidosis but without multiple myeloma had high ratios;
in three the ratio was greater than 16. The immunoperoxidase technic does not differentiate amyloidosis from
Vol. 85 • No. 6
693
IMMUNOPEROXIDASE TECHNIC
multiple myeloma but may be helpful in distinguishing
primary amyloidosis from monoclonal gammopathy. of
undetermined significance.
Three patients with renal disease secondary to monoclonal proteins also were studied. One of the three patients
had light chain nephropathy based on a renal biopsy and
no other evidence of multiple myeloma. In all three patients the light chain ratios were markedly abnormal. Since
these patients are sometimes difficult to diagnose, especially in the absence of a renal biopsy, determination of
light chain ratios could prove useful.
In summary, this communication confirms previous
reports5'12 of the feasibility and specificity of the immunoperoxidase method as applied to bone marrow biopsies.
Furthermore, the study defines a light chain ratio (>16)
that may aid in the differentiation of patients with multiple
myeloma from those with monoclonal gammopathy of
undetermined significance. It further suggests that this
high ratio, when found in patients without overt myeloma,
may indicate the presence of a disease process in which
monoclonal proteins cause end organ damage, such as
that occurring in amyloidosis or in renal disease secondary
to a monoclonal protein. Further studies are needed to
determine if this technic may be helpful in the clinical
follow-up of patients with multiple myeloma and in the
identification of patients with monoclonal gammopathy
who progress to an overt plasma cell malignancy.
Acknowledgment. The authors thank Dr. Robert Sherman for assisting
with the statistical analysis.
References
1. Axelsson U, Bachmann R, Hallen J: Frequency of pathologic proteins
(M-components) in 6995 sera from an adult population. Acta
MedScand 1966; 179:235-247
2. Brynes RK, McKenna RW, Sundberg D: Bone marrow aspiration
and trephine biopsy. Am J Clin Pathol 1978; 70:753-759
3. Greipp PR, Kyle RA: Clinical, morphological, and cell kinetic differences among multiple myeloma, monoclonal gammopathy of
undetermined significance, and smoldering multiple myeloma.
Blood 1983;62:166-171
4. Hallen J: Discrete gammaglobulin (M-) components in serum. Acta
Med Scand (Suppl) 1966; 462:1 -127
5. Hitzmanri JL, Chin-Yang L, Kyle RA: Immunoperoxidase staining
of bone marrow sections. Cancer 1981; 48:2438-2446
6. Hsu S, Raine L, Fanger H: The use of antiavidin antibody and avidinbiotin-peroxidase-complex in immunoperoxidase technics. Am
J Clin Pathol 1981;75:816-821
7. Jamshidi K, Swaim WR: Bone marrow biopsy with unaltered architecture: A new biopsy device. J Lab Clin Med 1971; 77:335342
8. Kyle RA: Monoclonal gammopathy of undetermined significance.
Natural history in 241 cases. Am J Med 1978; 64:814-826
9. Kyle RA, Greipp PR: Smoldering multiple myeloma. N Engl J Med
1980;302:1347-1349
10. Maldonado JE, Velosa J A, Kyle RA, Wagoner RD, Holley KE,
Salassa RM: Fanconi syndrome in adults. A manifestation of a
latent form of myeloma. Am J Med 1975; 58:354-364
11. Martelli MF, Falini B, Tabilio A: Multiple myeloma and related
conditions: An immunoperoxidase study of paraffin-embedded
bone marrow biopsies. Haematologica 1982; 67:1-16
12. Pinkus GS, Said JW: Specific identification of intracellular immunoglobulin in paraffin sections of multiple myeloma and macroglobulinemia using an immunoperoxidase technique. Am J
Pathol 1977; 87:47-55
13. Taylor CR, Russell R, Chandor S: An immunohistologic study of
multiple myeloma and related conditions, using an immunoperoxidase method. Am J Clin Pathol 1978; 70:612-622