CLINICAL MICROBIOLOGY
Original Article
Evaluation of Commercial Enzyme
Immunoassays Compared to
Immunofluorescence and Double Diffusion
for Autoantibodies Associated with
Autoimmune Diseases
KAREN JAMES,
PH.D.,
AND GAYLE MEEK, M.T.(ASCP)
A commercially available enzyme immunoassay system for detecting autoantibodies to double-stranded DNA, deoxyribonuceloprotein, Smith, ribonuclearprotein, Sjogren's syndrome-associated antigens A and B, and scleroderma-associated antigen
70 was compared to the conventional immunofluorescence assay
for double-stranded DNA and double diffusion assays for extractable nuclear antigens. There was excellent correlation between methods, but it appears that the enzyme immunoassays
are more sensitive. Based on the results of this study, the authors
recommend performing anti-nuclear antibody screening at two
dilutions, with enzyme immunoassay follow-up of appropriate
patient sera that are positive on anti-nuclear antibody testing.
Nucleolar and centromere pattern anti-nuclear antibodies are
diagnostic for variants of scleroderma and need no further evaluation. Negative anti-nuclear antibody tests performed using
HEp-2 tissue culture cells require no further evaluation. (Key
words: Collagen-vascular diseases; Enzyme immunoassay; Antibodies to extractable nuclear antigens; Antibodies to DNA;
Autoantibodies) Am J Clin Pathol 1992;97:559-565
Detecting specific autoantibodies is useful when diagnosing and monitoring treatment of patients with various
collagen-vascular diseases, e.g., systemic lupus erythematosus (SLE), Sjogren's syndrome, progressive systemic
sclerosis, and mixed connective tissue disease. The methods most frequently used in the laboratory are indirect
fluorescent assay (IFA) for anti-DNA and double diffusion
(DD) for antibodies to extractable nuclear antigens (ENA).
Indirect fluorescent assay requires several hours of assay
time; reading fluorescent tests is subjective and dependent
on training of the technologists and on the specificity of
the substrate. Double diffusion requires 24 to 48 hours of
incubation, and reading the lines of identity, partial identity, and nonidentity is subjective and requires well-trained
and experienced technical staff.
Enzyme immunoassay (EIA) techniques are now commercially available to detect these specific autoantibodies.
Enzyme immunoassay technology is an objective measurement of the presence or absence of autoantibodies.
In addition, EIA provides a quantitative measurement
that may be useful in monitoring therapy in certain patients. This study was designed to evaluated the feasibility
and accuracy of EIAs to detect autoantibodies in a community hospital setting.
MATERIALS AND METHODS
Sera
Extra sera from 105 patient specimens submitted to
the community hospital laboratory for anti-nuclear antibody (ANA) evaluation were used. Sera selected for testFrom the Department of Anatomic and Clinical Pathology, Central
ing in this study included 27 with high-titered (> 1:320)
DuPage Hospital. Winfield, Illinois.
diffuse (homogenous) ANAs, 24 with high-titered
Received May 28, 1991; received revised manuscript and accepted for (> 1:160) speckled ANAs, 9 with low-titered (< 1:80)
publication September 19, 1991.
speckled ANAs, 10 with nucleolar ANAs, 11 with mixed
Address reprint requests to Dr. James: Department of Anatomic and
and/or
unusual ANA patterns, 20 with negative ANAs
Clinical Pathology, Central DuPage Hospital, 25 North Winfield Road,
Winfield, Illinois 60190.
(screened at 1:40), 2 with negative ANAs but positive anti559
560
CLINICAL MICROBIOLOGY AND IMMUNOLOGY
/ Article
mitochondrial antibodies (AMA), and 2 with negative
ANAs but positive anti-smooth muscle antibodies. Before
testing, all sera were initially thawed, diluted as specified
for the procedure, divided into aliquot portions, and refrozen at - 7 0 °C until they were reassayed.
Clinical Correlations
Chart review was performed for all patients with positive results for any of the assay systems. In cases in which
the hospital record did not clearly specify the criteria used
to determine the diagnosis, the physicians were contacted
personally.
Anti-Nuclear
Antibodies
Anti-nuclear antibodies were evaluated by indirect immunofluorescence using kits purchased from ImmunoConcepts Inc. (Sacramento, CA). These kits use HEp-2
cells as the substrate and fluorescein-conjugated anti-immunoglobulin (Ig) G (heavy and light chain) to detect the
antibodies in patient sera. The procedure specified in the
package insert was followed precisely.
Anti-DNA by Indirect
Immunofluorescence
Anti-DNAs were detected and quantitated using kits
purchased from Kallestad Laboratories, Inc. (Chaska,
MN). These kits use Crithidia luciliae as the substrate and
fluorescein-conjugated antisera with specificity for IgG,
IgA, and IgM (heavy and light chain) to detect the antibodies in patient sera. The procedure specified in the
package insert was followed precisely to detect anti-double-stranded (ds) DNA antibodies by IFA. For evaluation
of possible false-positive antibodies to DNA in C. luciliae,
histone was extracted from the organisms by 0.1 mol/L
HC1, as previously described.' Controls of known positive
and negative antibody reactivity for dsDNA were included
in these additional studies to insure the retention of
dsDNA reactivity.
Ouchterlony Double Diffusion for
Autoantibodies
Two slightly differing methods were used to detect antibodies by Ouchterlony DD to the following antigens:
Smith (Sm), ribonucleoprotein (RNP), Sjogren's syndrome-associated antigens A and B (SS-A and SS-B), and
scleroderma-associated antigen (Scl-70). All samples were
evaluated using the Auto I.D. Autoantibody Test System
<
FIGS. 1A and B. Pattern of sera applied for autoantibody identification
using DD. (A.) To screen patient samples, the following substances are
placed into wells numbered: (1) antibody-positive control; (2-6) up to
five patient sera; (7) (center well) polyvalent nuclear antigen. To identify
specific antibody reactions in patient samples that were positive in the
screening process, the following substances are placed into wells numbered: (1 and 4) two patient sera; (2) Sm-positive antibody control; (3)
RNP-positive antibody control; (5) SS-A-positive antibody control; (6)
SS-B-positive antibody control; (7) (center well) polyvalent nuclear antigen. Note: Scl-70 or Jo-1-positive antibody control could be used in
place of other positive controls if the patient sera reacts with the nuclear
antigen but does not show an identity reaction with any of the more
common known antibodies. (B.) Each patient serum must be evaluated
separately for specific antibodies in three different test systems unless
the antibody specificity is known. The following substances are placed
into wells numbered: (1) antibody-positive control (monospecific); (2)
patient #1 (~80 ML); (3) patient #1 (~20 nL); (4) antibody positive
control (monospecific); (5) patient #2 (~20 nL); (6) patient #2 (~80
tiL); (7) antigen (one of three, depending on specificity): monovalent
Scl-70 or divalent SS-A and SS-B or divalent Sm and RNP.
A.J.C.P. • April 1992
JAMES AND MEEK
Commercial EIAs Compared to Immunofluorescence and DD for Autoantibodies
(purchased from ImmunoConcepts, Inc., Sacramento,
CA), and 30 samples also were tested with the INOVA
Autoantibody Test System (provided by INOVA Diagnostics, Inc., San Diego, CA).
The Auto I.D. DD test is performed routinely in our
laboratory as specified in the product insert. The central
well of the supplied agarose plate was filled with the multispecificity nuclear antigen (20 ML). The antigen for this
test system contains 7 known (Sm, RNP, SS-A, SS-B, Scl70, proliferating cell nuclear antigen (PCNA), Jo-1) and
an undetermined number of unknown extractable nuclear
antigens. Patient samples and control specimens (20 fiL)
were placed into the six wells surrounding the antigen
well (Fig. \A). The plates were incubated at room temperature for 18 to 24 hours and read using a backlighted,
magnified viewing box (Kallestad Laboratories, Inc.).
Samples that appeared negative after 24 hours were incubated an additional 24 hours and read again. Patient
specimens that tested positive (a visible line of precipitation) were retested using alternate wells filled with control sera containing antibodies of known specificities (Fig.
\A). Reactions of identity, partial identity, or nonidentity
were read and interpreted for each positive patient sample.
Patient sera were diluted > 1:2 using phosphate-buffered
saline with 10% EDTA to chelate the calcium to prevent
nonspecific precipitin reactions between C-reactive protein and the nuclear antigen extract (unpublished observation), and applied twice to each well. In certain patient
samples with weakly reactive antibodies, the antigen was
diluted 1:2 and reincubated with undiluted patient sera
to define more clearly the identity or partial identity reactions.
The INOVA DD testing was performed on selected patients using limited materials supplied by the manufacturer for this study. Patient specimens were selected for
testing on the INOVA system to confirm positive results
and to reevaluate contradictory results obtained between
the Auto I.D. test system and the EIA methods described
561
below. The configuration of the INOVA method is unique
(Fig. IB). Each patient sample is placed in a large upper
well (approximately 80 ^iL) and a small lower well (approximately 20 ^iL). Specific nuclear antigens (Sm/RNP
separate from SS-A/SS-B separate from Scl-70) are placed
into the center well (approximately 80 fiL). The two different volumes of patient sera applied for each test eliminate the need for dilutions and theoretically decrease the
likelihood of a prozone reaction. Controls appropriate for
the specific nuclear antigen(s) being tested were placed in
the specified outer wells. Because the INOVA method
uses three separate antigens, each patient sample required
testing on multiple plates. The INOVA plates were incubated and interpreted as described above for the
ImmunoConcepts DD method.
EIA Using Antigen-Coated Microtiter Trays
(Microassay)
These products were provided by Diamedix Corporation (Miami, FL). Seven microassays were used to
detect antibodies to the following antigens: dsDNA, deoxyribonucleoprotein (DNP), Sm, RNP, SS-A, SS-B, and
Scl-70. Each antigen is bound to a separate strip of microtiter wells that facilitates testing groups of seven patient
sera and controls for all seven antigens simultaneously.
If fewer than seven patients are to be tested, unused wells
can be removed from the strip and used at a later time.
Diluted patient sera were added to the wells and incubated at ambient temperature for 30 minutes. After
washing to remove unbound specimen, the alkaline phosphatase-labeled conjugate (anti-human IgG and IgM
[DNA and DNP] or anti-human IgG [all antibodies to
ENAs], with heavy and light chain reactivity) was added
and incubated at room temperature for 30 minutes. The
unbound conjugate was removed by washing and the enzyme substrate (p-nitrophenyl phosphate) was added. The
enzymatic reaction was stopped after 15 or 30 minutes
with 0.5 mol/L trisodium phosphate to fix the intensity
TABLE 1. EIA MICROASSAY RESULTS CATEGORIZED BY ANA PATTERN
ANA Staining Pattern
Microassay (EIA)
Anti-DNA
Anti-DNP
Anti-Sm
Anti-RNP
Anli-SS-A
Anti-SS-B
Anti-Scl70
Diffuse
> 1:320
n = 27
Speckled
n = 24
Low Speckled
<.1:80
n = 9
Nucleolar
n = 10
Mixed
n = 11
Negative
n = 24
9(4)
18
2
3
4
2
2
2(3)
2
1
2
3
2
2
0
0
0
0
0
1
0
0(1)
0
0
0
0
0
0
4
5
0
0
3
1
2
0(1)
0
0
0
0
1
0(1)
2.1:160
The number of positive patient specimens (borderline patient specimens) for the microassay
autoantibody EIA for each grouping of ANA staining patterns. The patient sera positive for AMA
and ASMA were negative by microassay and were included with the ANA-negative data.
Vol. 97. No. 4
CLINICAL MICROBIOLOGY AND IMMUNOLOGY
562
Original Article
of the color. The intensity of the reaction, which was read
photometrically at 405 nm using the BP-12 strip reader
(Diamedix Corporation, Miami, FL), is directly proportional to the amount of patient autoantibody in the specimen.
Two dilutions of patient sera were prepared. A 1:41
dilution (used for the DNA and DNP assays) was prepared
by adding 10 /iL of patient sample to 400 /xL of diluent;
and a 1:101 dilution (used for the SS-A, SS-B, Sm, RNP,
and Scl-70 assays) was prepared by adding 5 nL of patient
sample to 500 nL of diluent. Although there are two dilutions of patient serum to prepare, the rest of the assay
procedure is identical for all antigens. The manufacturer's
instructions were followed precisely. Specimens positive
for each EIA were retested using 12 replicate wells of each
separate EIA plate to determine intraassay reproducibility.
Results are reported in either enzyme-linked immunosorbent assay units (EU) per milliliter or in International Units (IU) per milliliter when IU standards have
been established. The respective units were calculated by
comparison to a calibrator that was assayed with each
batch of tests. Positive, negative, and borderline result
ranges have been established so patient results can be determined from the single-point calibrator. The interpretations used were slightly different than those recommended by the manufacturer at the time of this study:
DNA: < 120 IU/mL = negative, 121-300 IU/mL = borderline, > 300 IU/mL = positive; DNP: < 40 IU/mL
= negative, > 40 IU/mL = positive; SS-A, SS-B, Sm,
RNP, Scl-70: < 25 EU/mL = negative, > 25 EU/mL
= positive.
RESULTS
The microassay for autoantibodies was highly reproducible. Coefficients of variation were DNA = 5.05, DNP
= 2.87, Sm = 0.90, RNP =1.81 SS-A =1.75, SS-B = 1.27,
and Scl-70 = 0.77. Of the 110 specimens originally tested,
five were from the same patients at different times. The
positive, borderline, or negative interpretations of all five
specimens were repeatable despite intervals of S; 3 months
between sampling, e.g., a patient with SLE had DNAs of
1,134 and 1,486 IU/mL, DNPs of 422 and 552, Sms of
87 and 83, RNPs of 66 and 25, SS-As of 132 and 180,
SS-Bs of 5 and 9, and Scl-70s of 21 and 12. The second
specimen for each of the duplicated patients was deleted
from the study to report data from 105 unique patients.
As expected, specimens with diffuse ANA patterns
yielded the greatest number of positive anti-DNA and
anti-DNP results by microassay (Table 1). Three other
autoantibodies (Sm, RNP, SS-A) were detectable in three
patients with classic SLE, the disease most often associated
with multiple autoantibodies. 23 Two patients with diffuse
ANA patterns had anti-DNA, anti-DNP, and anti-RNP.
Two patients with diffuse ANA patterns had anti-Scl-70
by all methods; the reevaluated ANA patterns were actually nucleolar and nucleoplasm^, the classic ANA pattern associated with the Scl-70 antigen as described by
Tan and colleagues.4
Two patients with high-titered speckled-pattern ANAs
had strongly positive anti-DNA by IFA as well as by EIA;
both patients had definitive diagnoses of SLE. Another
lupus patient had a borderline anti-DNA by EIA, a positive anti-DNA by IFA, and positive antibodies to DNP,
RNP, and SS-A. One patient with a low-titered speckledpattern ANA, a detectable anti-SS-B, and clinical symptoms consistent with Sjogren's syndrome.
The patient sera classified as mixed-pattern ANAs were
a very heterogeneous group and serve to illustrate the value
of performing specific autoantibody detection. Four pa-,
tients had the double pattern of diffuse and nucleolar:
only one was positive for Scl-70, one was positive for SSA and SS-B, two were positive for anti-DNA and antiDNP, and one patient was positive for anti-SS-A alone.
There were four patients with the double patterns of
TABLE 2. ANTI-DNA BY IFA AND ENA BY OUCHTERLONY CATEGORIZED BY ANA PATTERN
ANA Staining Pattern
DNA by IFA
or
ENA by DD
Anti-DNA
Anti-Sm
Anti-RNP
Anti-SS-A
Anti-SS-B
Anti-Scl70
Diffuse
> 1:320
Speckled
^1:160
Low
Speckled
<1:80
Nucleolar
Mixed
Negative
10
2
3
3
1
2
3
1
2
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
0
3
1
1
0
0
0
0
1
0
The number of positive patieni specimens in the indirect immunofluorescence assay (IFA) for
anti-DNA or the number of positive patient specimens in the double diffusion (DD) method for
ENA for each grouping of ANA staining patterns. For the DD assays, patient specimens were
positive in either the INOVA method, or the ImmunoConcepts method, or both to be included
in the data presented above. As in Table I. the data for patients with AMA and ASMA arc
included with the ANA-negative data.
A.J.C.P.-April 1992
JAMES AND MEEK
Commercial EIAs Compared to Immunofluorescence and DD for Autoantibodies
diffuse and speckled; one well-characterized SLE patient
had multiple autoantibodies, including antibodies to
DNA, DNP, SS-A, and a strongly positive Scl-70. Two
remaining patients had low-titered diffuse and speckled
mixed patterns; one was negative for all tested autoantibodies and one had a repeatedly borderline anti-DNA.
Table 2 shows the distribution of results by the "standard" methods of anti-DNA by IFA and anti-ENAs by
DD. Results were very similar to those shown in Table 1
and differences are attributable to more sensitive results
in the microassay. The two patients' sera that contained
autoantibodies detectable by DD but not by EIA (Table
3) were confirmed SLE patients. One patient had very
high Sm antibodies but negative RNP antibodies by EIA,
whereas both DD methods showed a "spur" reaction of
partial identity, characteristic of both Sm and RNP antibodies (Table 3B). The other lupus patient had Scl-70
by both DD methods but was negative by EIA (Table 3E).
No other instances of false-negative EIA results were detectable.
There were 13 occurrences of positive EIA results with
negative DD assays (Table 3). DD- and EIA-positive results occurred with all anti-Sm (Table 3A) or with antiRNP (Table 3B). Of the four patients with positive EIAs
(but negative DD) for anti-SS-A, two were well-diagnosed
SLE patients. Another patient whose SS-A was 38 EU/
mL had a clinical diagnosis of SLE by American Rheumatism Association criteria.5 The fourth patient had a
very low level (25 EU/mL) of anti-SS-A as well as antiDNA and anti-DNP; this patient had procainamide-induced lupus with significant improvement in symptoms
after the drug was discontinued, but a diagnosis of SLE
has not been confirmed.
Four patient sera contained antibodies to SS-B by EIA
but not by DD. All four patients had low positive levels
(31, 26, 31, and 49 EU/mL) that on repeated testing were
TABLE 3. SPECIFIC ENA EIA MICROASSAY RESULTS
COMPARED TO DOUBLE DIFFUSION RESULTS
Positive
A. Diffusion
Positive
Negative
B. Double Diffusion
Positive
Negative
C. Double Diffusion
Positive
Negative
D. Double Diffusion
Positive
Negative
E. Double Diffusion
Positive
Negative
Negative
Anti-Sm Microassay
3
0
0
102
Anti-RNP Microassay
5
1
0
99
Anti-SS-A Microassay
6
0
4
95
Anti-SS-B Microassay
3
0
4
98
Scl-70 Anti-Microassay
2
1
5
97
563
TABLE 4. ANTI-DNA BY EIA MICROASSAY COMPARED
TO ANTI-DNA BY IFA
Anti-DNA by Microassay
Anti-DNA
by IFA
Positive
Borderline
Negative
Positive
Negative
12
3
6
2
0
82
still positive (27, 29, 52, and 41 EU/mL, respectively).
Only one of these patients (49 and 41 EU/mL) had documented symptoms that would be consistent with Sjogren's syndrome. Three of five patients with Scl-70 antibodies by EIA but not by DD are the same lupus patients
who had multiple autoantibodies.
The DNA assay by EIA showed some significant differences from the C. luciliae IFA assay. Four patients were
negative on the EIA and positive for anti-DNA by IFA;
two patients had anti-DNP, another had anti-SS-A, and
the fourth had anti-DNP and anti-RNP. Three patients
were anti-DNA negative by IFA and positive by EIA (Table 4).
Although both assay systems for anti-DNA (IFA and
EIA) use antisera with specificity for IgM and IgG heavy
and light chains (the IFA antisera also contains specificity
to IgA), the sensitivity of each assay system for IgG and
IgM heavy chains may be different. Separate and specific
IgM and IgG assays were performed to determine the effect
of IgM compared to IgG for three false-negative and three
false-positive sera for anti-DNA by EIA and compared to
anti-DNA by IFA. Only one patient's false-positive (borderline) EIA result could be explained by IgM antibodies.
IgM antibodies did not explain any of the false-negative
antibodies to DNA measured by EIA.
One of the four patients with a false-negative result for
anti-DNA by EIA had rheumatoid factor, but the other
three were negative, so rheumatoid factor did not explain
the differences in results. When HCl-extracted C. luciliae
were used, however, all four patient sera became negative
for anti-DNA by IFA.1
Only one of the false-positive EIAs was significantly
and repeatedly higher (890 and 740 IU/mL) than the borderline range. This patient also tested positive using another commercial EIA for anti-dsDNA (INOVA Diagnostics), had a positive anti-DNP, and clinically had procainamide-induced lupus.
There were only two false-negative EIA ENAs, and both
were found in SLE patients with multiple other autoantibodies. This suggests that reading DD reactions is more
difficult than interpreting EIA results. Four positive EIA
results (but negative DD) were low levels of SS-B that
could be considered "borderline" and may be early predictors of Sjogren's syndrome.6 Five positive EIA results
Vol. 97 • No. 4
CLINICAL MICROBIOLOGY AND IMMUNOLOGY
564
Original Article
were found in three SLE patients with multiple antibodies
that perhaps were not discernible from other antibodies
on DD. 7 Two patients with definite levels of SS-B and
Scl-70 by EIA, not detectable by double diffusion, were
undoubtedly early cases of Sjogren's syndrome and progressive systemic sclerosis, respectively, based on their
clinical presentation at the time of testing.
DISCUSSION
The most significant finding of this study is that specific
autoantibodies cannot be predicted reliably by the ANA
pattern. Reading of ANAs is subjective and ANA patterns
can be read differently, as evidenced by the 1990 S-D
Immunology Survey of the College of American Pathologists proficiency testing program. In that survey, the same
specimen (S-68) was reported as diffuse (homogeneous)
by 82% of the participants, speckled by 12.5%, and "more
than one pattern present" by 5.5%, regardless of substrate
used.8 As shown in this study, several patients with diffuse
ANAs had positive ENA antibodies and two patients with
unquestionably speckled ANAs had strongly positive antibodies to DNA. The subjectivity of reading ANAs as
well as the inter- and intralaboratory variability strongly
suggest that there is a need for further evaluation of all
definitely positive ANAs by a standard panel of well-characterized antigens (DNA and ENA antigens).
Borderline or questionable ranges need to be established
for EIAs because cut-off points are arbitrary and comparison with IFA or DD always results in some discrepancies between the methods.6 Enzyme immunoassay results in a borderline range should be repeated and verified
before reporting.
Negative ANAs do not require further evaluation. In
fact, requests to perform further testing on ANA-negative
patients should be denied if the laboratory is using tissue
culture substrates (HEp-2 or Wil-2 human cell lines). Animal substrates (rat or mouse liver or kidney), however,
may not detect SS-A, SS-B, or centromere antibodies.3,4
Low-titered, speckled ANAs may not require further testing, but that proposal needs further study with more patients.
The detection of DNA antibodies is the most troublesome. There has not been a perfect method to test for
anti-DNA. The Fair assay had false-positive results due
to single-stranded DNA. The C. luciliae assay detects only
dsDNA, but it also detects antibodies to histone as shown
in this report and by others.'-7 Although the number of
specimens evaluated was limited, the microassay reported
here identified all 18 patients who definitely had DNA
antibodies. Laboratories using the C. luciliae assay obviously have been reporting false-positive anti-DNAs,' 7
but there also may be an occasional false-positive result
(based on clinical criteria) with the microassay, as described in this report.
The microassay system described here uses ENA antigens purified from calf thymus nuclei. The study described is the first to compare the commercially available
purified ENA antigens in a community hospital setting.
A test strip format using recombinant antigens is commercially available and has been evaluated by a reference
laboratory.7
The advantages of the EIA microassay for detecting
autoantibodies include (1) that EIA is objective, yielding
numeric values; (2) increased sensitivity for detecting ENA
antibodies; (3) positive results are quantitative without
retesting dilutions; (4) nonspecific reactivities attributable
to C-reactive protein or other proteins that precipitate
with agarose are eliminated; (5) subjective reading of
identity, partial identity, and nonidentity reactions is
eliminated; and (6) low-titered antibodies are not obscured
by high-titered antibodies. Disadvantages of an EIA system for detecting autoantibodies include (1) that not all
ANA specificities are characterized and some might be
missed by using the available specific antigens in an EIA
and (2) increased sensitivity and specificity could be con-
TABLE 5. PROPOSED REFLEX TESTING AND INTERPRETIVE REPORTING BASED ON ANA PATTERN
ANA Result
Interpretive
Reflex Testing
Negative
None
> 1:40 < 1:160, Patterns:
speckled, diffuse
None
> 1:160, Patterns: speckled,
diffuse, peripheral, or mixed
a 1:40 Nucleolar
£ 1:40 Centromere
EIA microassay or
anti-DNA and ENA
None
None
S:l:40 Mixed pattern
EIA microassay or
anti-DNA and ENA
Reporting
Absence of an ANA would rule out collagen vascular autoimmune
disease at this time.
Low titered ANAs of these patterns are not diagnostic of collagen
vascular autoimmune disease at this time. Suggest reevaluation in
6 to 12 months, if clinically indicated.
Significantly positive ANA titer and pattern. Suggest ENA/DNA
antibody panel to characterize further specificity of this ANA.
Nucleolar pattern ANAs are highly associated with scleroderma.
Centromere pattern ANAs are highly associated with CREST
syndrome.
Possibly significant ANA pattern; Suggest ENA/DNA antibody panel
to characterize further specificity of this ANA.
A.J.C.P.-April 1992
JAMES AND MEEK
Commercial EIAs Compared to Immunofluorescence and DD for Autoantibodies
fusing to clinicians until long-term studies are reported
and the clinical relevance of borderline positive results is
determined.
The EIA microassay to detect autoantibodies can be
substituted for DD and IFA methods. The EIA is more
accurate than the IFA assay to detect antibodies to
dsDNA. The EIA is more sensitive than the DD assays
for ENA antibodies.
The EIAs are more cost effective ($30.66 per test for
five ENA antigens or $43.23 per test for five ENA antigens
plus anti-DNA and anti-DNP), whereas the average IFA
costs $ 14.03 per test and the average DD costs $42.06 per
test. The cost accounting for all assays was based on testing
an average of four patients per assay. The average cost
per test was calculated based on a rate of positive antiDNAs of 21.5% (with positive tests subsequently titered);
the ENA average cost was based on a positive rate of 30.3%
(with positive tests subsequently reevaluated for identity
with known antibodies and titered).
Last, we propose (Table 5) that all requests for ANA
initially should be evaluated at two dilutions, 1:40 and
1:160, using HEp-2 cells. Sera with a definitely positive
(S; 2+) reaction at 1:160 dilution with peripheral, diffuse,
speckled, or mixed patterns should be evaluated further
with the microassay system or by both anti-DNA by IFA
and ENA antibodies by DD. Reports of patient sera with
these patterns of reactivity at 1:40 dilution and weak or
negative reactions at 1:160 dilution should be accompanied by this interpretive comment: "This ANA titer may
be nondiagnostic at this time. Suggest reevaluation in 6
565
months." Nucleolar and/or centromere antibodies are diagnostic at any titer without any further evaluation. Patients with negative ANAs need no further evaluation for
autoantibodies. If, however, symptoms persist, another
ANA at a future date would be indicated. This proposal
would expedite the cost-effective evaluation of patient sera
for autoantibodies.
REFERENCES
1. Deng JS, Rubin RL, Lipscomb MF, Sontheimer RD, Gilliam JN.
Reappraisal of the specificity of the Crilhidia luciliae assay for
nDNA antibodies: Evidence for histone antibody kinetoplast
binding. Am J Clin Pathol 1984;82:448-452.
2. Boey ML, Peebles CL, Tsay G, Feng PH, Tan EM. Clinical and
autoantibody correlations in Orientals with SLE. Ann Rheum
Dis 1988;47:918-923.
3. Nakamura RM, Tan EM. Repent advances in laboratory tests and
the significance of autoantibodies to nuclear antigens in systemic
rheumatic disease. Clin Lab Med 1986;6:41-53.
4. Tan EM. Autoantibodies to nuclear antigens (ANA): Their immunobiology and medicine. Adv Immunol 1982;33:167-240.
5. Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for
the classification of systemic lupus erythematosus. Arthritis
Rheum 1982;25:1271-1277.
6. Venables PJW, Smith PR, Maini RN. Purification and characterization of Sjogren's syndrome A and B antigens. Clin Exp Immunol 1983;54:731-740.
7. Paxton H, Bendele T, O'Connor L, Haynes DC. Evaluation of the
RheumaStrip ANA profile test: A Rapid test strip procedure for
simultaneously determining antibodies to autoantigens Ul-ribonucleoprotein (Ul-RNP), Sm, SS-A/Ro, SS-B/La, and to native
DNA. Clin Chem 1990;36:792-797.
8. College of American Pathologists Diagnostic Immunology Series 2
Survey Summary Data Reference for Set S-D, 1990.
Vol. 97 • No. 4