Serological tests for diagnosis of systemic lupus

Commentary and update: Serological tests for diagnosis of systemic lupus erythematosus (SLE)

John D. Clough, M.D. Leonard H. Calabrese, D.O.

Since the description of the lupus erythemato-sis (LE) cell phenomenon by Hargraves et al,1

and with the realization by Haserick and Bortz2

that this phenomenon could be induced in nor-mal cells by a component of plasma f rom patients with systemic lupus erythematosus (SLE), the na-ture of this "LE cell factor" has been clarified. It is an IgG antibody that reacts with deoxyribonu-cleoprotein3 and as such belongs to an increas-ingly well-characterized group of antibodies that combine with various components of cell nuclei. Detection of certain of these antinuclear antibod-ies (ANA) has assumed increasing importance in the diagnosis and management of SLE and re-lated diseases.

Among the serological tests for diagnosis of SLE, the LE cell phenomenon was followed by fluorescent assays for ANA with the use of var-ious cellular substrates,4 5 tests for antibodies against double- or single-stranded DNA, 6 - 8 and assays for antibodies against an acidic nuclear glycoprotein refer red to as Sm antigen.9 Al-though these assays are in widespread use, and in many institutions combinations of them are of ten ordered as a "lupus battery," little quantitative information about the meaning and interpreta-tion of positive or negative test results is available in the literature. We reviewed the laboratory and clinical records of T h e Cleveland Clinic Foun-dation, including Haserick's original work to up-

1 D e p a r t m e n t of Rheuma t i c and Immuno log ic Disease, a n d the D e p a r t m e n t of I m m u n o p a t h o l o g y , T h e Cleveland Clinic F o u n d a -

date and evaluate serological tests for diagnosis of SLE.

Methods

Antinuclear antibody (ANA) was assayed by indirect immunofluorescence5 with the use of rat kidney as substrate and a polyvalent, fluorescein-ated rabbit anti-human immunoglobulin serum as indicator; a titer > 1:40 was considered posi-tive. Anti-native DNA (anti-nDNA) was assayed by a modification of the Farr assay as previously described10; values > 10% binding were consid-ered positive. Anti-Sm was assayed by double diffusion in agarose with the use of ribonuclease-treated rabbit thymus ex t rac t " ; positive sera formed a line of identity with a known positive standard (kindly supplied by Carol Peebles, Na-tional Jewish Hospital, Denver, Colorado). T h e LE cell test was pe r fo rmed as previously de-scribed by Haserick and Bortz2; positives were ranked 1 -I- or greater.

Sensitivities (prevalence of true-positives in the SLE population) were determined by performing each of the tests on sera f rom a group of 102 well-studied SLE patients previously reported.1 1

Specificities (prevalence of true-negatives in the non-SLE population) were determined by review-ing clinical records of 573 patients selected for test positivity or negativity in approximately equal numbers for each test. These patients were classified as SLE or non-SLE; f rom this informa-tion together with the frequency of overall test positivity, spécificités were calculated. In most cases data f rom more than one test were available

65

on December 30, 2021. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from

http://www.ccjm.org/

66 Cleveland Clinic Quar ter ly Vol. 50, No. 2

on each pat ient . Sensitivities a n d specificities a re shown in the Table.

Resu l t s

Predict ive values'" for positive and negative results of each of the fou r tests a re shown in Figures 1 and 2, respectively. T h e predict ive value of a positive test is the prevalence of SLE a m o n g all pat ients with a positive test result; the tests with the highest positive predict ive values were an t i -nDNA and anti-Sm. T h e predict ive value of a negat ive test result is the preva lence of non-SLE a m o n g all pat ients with a negat ive test result; the test with the highest negat ive predict ive value was A N A .

T e s t eff ic iency '" is de f ined as the prevalence of t rue-posi t ive and t rue-negat ive results a m o n g all the tests p e r f o r m e d . It measures the correla-tion of test results with clinical si tuation (sum of SLE pat ients with positive test and non-SLE pa-tients with negat ive test divided by total n u m b e r

Table. Sensitivity and specificity of serological tests for SLE

lest Sensitivitv Spenfidiv

LE Prep .704 .946 A N A .990 .690 Ant i -n I )NA .569 .993 Anti-Sm .220 .999

A N A = an t inuc lea r ant ibodies .

of tests done) . Efficiency for tests examined is shown in Figure 3; t h e most efficient test was the LE cell p repara t ion .

Quant i ta t ive corre la t ions were d rawn fo r A N A ti ter and d e g r e e of positivity of LE p repara t ion with prevalence of SLE as well as with certain o the r non-SLE diagnoses (Figs. 4 and 5, respec-tively). In both tests the prevalence of SLE in-creased as d e g r e e of test positivity increased, a l though this re la t ionship was c learer for LE cell p repara t ion .

Discuss ion

We e x a m i n e d the results of fou r tests (LE cell p repara t ion , A N A , an t i -nDNA, and anti-Sm) in 675 patients , 102 of whom were selected because they were diagnosed on clinical g r o u n d s as hav-ing SLE; the r e m a i n d e r were selected f r o m lab-ora tory records of the fou r tests. T h e results permit calculation of sensitivity and specificity figures (Table), of predic t ive values of positive and negat ive results (Figs. 1 and 2), and of test efficiency (Fig. 3) fo r each test. T h e s e f igures show how strongly a positive or negat ive test result s tanding alone, wi thout o the r da ta , indi-cates a diagnosis of SLE. T h e specificity figures in the Table a r e probably conservative (underes-t imate specificity) since the results are biased by the fact that the test was o r d e r e d in the first place; non-SLE pat ients on whom A N A was or-

100 co LxJ

+ 75 -Li_ O

UJ

< > 50

o

Q LLI tr O .

ANA LE anDNA aSm

100

c o Lil t 75

% 50 < > UJ

2 25 o o Lui tr Q-

ANA LE anDNA aSm F i g u r e 1. Predict ive value of positive results f o r 4 serological tests used in the diagnosis of SLE; value is the pe rcen t of pat ients with a

positive test who actually have SLE. F i g u r e 2. Predict ive value of negat ive results f o r 4 serological tests used in the diagnosis of SLE; value is the percent of pat ients with

a nega t ive lest who d o not have SI.E.



Summer 1983 Commenta ry 67

100

CO 75

° 50 > o z u

y 25 Li_ U_ LlJ

1

ANA LE anDNA aSm F i g u r e 3 . Ef f ic iency of 4 se ro logica l tests used in d iagnos i s of

SLF.; va lue is t h e p e r c e n t of p a t i e n t s in w h o m t h e o u t c o m e of t h e test a g r e e s wi th t h e d iagnos is .

dered , fo r example , may be m o r e likely to have positive A N A than non-SLE pat ients in whom this was no t even cons idered . However , the tests a re valuable as discr iminators only within the popula t ion on which they a r e p e r f o r m e d , so this bias is probably not especially impor tan t .

T h e results illustrate the principle that for a highly specific test (such as an t i -nDNA o r anti-Sm) the predict ive value of a positive resul t is high whereas for a highly sensitive test (such as ANA) the predict ive value of a negat ive test is

high. T h e LE cell p r e p was in te rmed ia te in spec-ificity and sensitivity, b u t as a single test had the greates t efficiency.

It is clear, however , tha t these test results are never considered in a vacuum. T h e tests a r e o r d e r e d because of suspicion based on clinical g r o u n d s that they may be positive, and the result ob ta ined exer ts e i ther a positive or negat ive ef-fect on the clinician's est imation of the likelihood that the pat ient has lupus. T h i s clinical est imation of likelihood can be quan t i t a t ed with a scoring system based on the Amer ican Rheumat i sm As-sociation (ARA) cri teria fo r the classification of SLE, according to the sensitivity a n d specificity of the cri teria fo r diagnosis of SLE as we have previously r e p o r t e d . " T h i s allows establ ishment of pre tes t probabil i ty of SLE in a given pat ient . If this probabil i ty is low, a positive o u t c o m e of a test with a high positive predic t ive value (e.g., an t i -nDNA, anti-Sm) would have the most pro-f o u n d effect on the diagnost ic l ikelihood. O n the o t h e r hand , for a patient with a high pretest probabil i ty of SLE, the grea tes t effect on proba-bility would be exe r t ed by a negat ive o u t c o m e of a test with a high negat ive predic t ive value (e.g., ANA) . Because of its nonspecifici ty, a positive A N A exer ts little effect on the likelihood of SLE, and , because of relatively low sensitivity, negat ive an t i -DNA and anti-Sm results do not greatly in-f luence the likelihood of SLE. T h u s , as o f ten happens in mild or inactive SLE, the combinat ion of positive A N A , negat ive an t i -nDNA, and neg-ative anti-Sm is not especially helpful . If this

100

1:40 1:80 1:160 ANA (Titer)

D r u g - i n d u c e d LE M i s c e l l a n e o u s

RA SLE

21:320

O z o <

D

> -u z LLJ 13 o

SLE

RA D r u g - i n d u c e d S L E M i s c e l l a n e o u s

F i g u r e 4 . P r ed i c t i ve va lue of posi t ive resu l t s of A N A at d i f f e r e n t t i te rs f o r id iopa th ic SLE , d r u g - i n d u c e d LE, r h e u m a t o i d a r t h r i t i s (RA), a n d all o t h e r d i agnoses (miscel laneous) .

F i g u r e 5. P red ic t ive va lue of d i f f e r e n t levels of posit ivity of LF. cell p r e p a r a t i o n f o r id iopa th ic SLE , d r u g - i n d u c e d LE, r h e u m a t o i d a r t h r i t i s (RA), a n d all o t h e r d i a g n o s e s (miscel laneous) .



6 8 Cleveland Clinic Quarterly Vol. 50, No. 2

combination of results occurs in an individual for whom the pretest probability of SLE is low (e.g., a pat ient with fibrositis), the clinician is simply faced with the unpleasant task of explaining a positive A N A to someone in whom SLE was not strongly suspected initially. This realization might affect the decision to o rde r A N A in this setting.

However , ant i -nDNA and possibly anti-Sm have values in addition to their impor tance in diagnosis. Ant i -nDNA levels correlate well with activity of SLE8 and, indeed, ant i -nDNA is though t to be an important pathogenic antibody in SLE, particularly with glomerulonephri t is .1 4 '1 ' Fu r the rmore , the combination of positive anti-n D N A and positive anti-Sm has in o u r experience been associated with a more severe fo rm of lupus in which diffuse, proliferative glomerulonephri t is is common (occurring in 73% of such patients) as opposed to patients in whom these antibodies are not found together , where severe renal disease is considerably less common (23%, p < 0.001).11

Thus , adequa te moni tor ing of disease activity requires serial determinat ions of ant i -nDNA along with certain o ther tests (complement, cre-atinine, urinalysis, blood count), and determina-tions of ant i -nDNA and anti-Sm may be helpful in estimating prognosis.

References 1. H a r g r a v e s M M , R i c h m o n d H , M o r t o n RJ. P re sen ta t ion of

two b o n e m a r r o w e lements : t h e " T a r t " cell a n d "LE" cell. P r o c Staff M e e t Mayo Clin 1948; 2 3 : 2 5 - 2 8 .

2. Hase r i ck J R , Bor tz D W . A new d iagnos t i c test f o r acu te d i s s emina t ed lupus e r y t h e m a t o s u s . Cleve Clin Q 1949; 16: 1 5 8 - 1 6 1 .

3. H o l m a n H R , D e i c h e r H R G , Kunke l H G . T h e L.E. cell a n d LE S e r u m fac tors . Bull N Y Acad M e d 1959; 3 5 : 4 0 9 - 4 1 8 .

4. H o l b o r o w t j , W e i r DM, J o h n s o n G D . A s e r u m f a c t o r in lupus e r y t h e m a t o s u s with a f f in i ty f o r t issue nucle i . Br M e d J 1957; 2 : 7 3 2 - 3 4 .

5. C l ey m ae t JE, N a k a m u r a R M . Ind i rec t i m m u n o f l u o r e s c e n t a n t i n u c l e a r a n t i b o d y tests: c o m p a r i s o n of sensitivity a n d spec-ificity of d i f f e r e n t subs t ra tes . A m J Clin Pa tho l 1972; 5 8 : 3 8 8 -393 .

6. Casals SP, Er iou GJ , Myers LL . S ign i f icance of an t i body to D N A in systemic l u p u s e r y t h e m a t o s u s . A r t h r i t i s R h e u m 1964; 7 : 3 7 9 - 3 9 0 .

7. A a r d e n L A , d e G r o o t ER, F e l t k a m p T E W . I m m u n o l o g y of D N A . I I I . Crithidia luciliae, a s imple s u b s t r a t e f o r t h e de t e r -mina t ion of a n t i - d s D N A with the i m m u n o f l u o r e s c e n c e tech-n ique . A n n NY A c a d Sei 1975; 2 5 4 : 5 0 5 - 5 1 5 .

8. C l o u g h j D . M e a s u r e m e n t of D N A - b i n d i n g i m m u n o g l o b u l i n s in systemic lupus e r y t h e m a t o s u s . J I m m u n o l M e t h o d s 1977; 1 5 : 3 8 3 - 3 9 4 .

9. T a n EM, Kunke l H G . Charac te r i s t i cs of a so luble nuc l ea r an t i gen p r e c i p i t a t i n g with sera of pa t i en t s with systemic lupus e r y t h e m a t o s u s . J I m m u n o l 1966; 9 6 : 4 6 4 - 4 7 1 .

10. K r a k a u e r RS, C l o u g h j D , A l e x a n d e r T , S u n d e e n J , S a u d e r D N . S u p p r e s s o r ce l l ,defect in SLE: re la t ionsh ip t o na t ive D N A b i n d i n g . Clin E x p I m m u n o l 1980; 4 0 : 7 2 - 7 6 .

11. C o u r i J M , Mowaf i N , C l o u g h j D . Associat ion of an t i -Sm with inc reased risk o f d i f f u s e lupus nephr i t i s (abst). Clin Res 1981; 29 :781 A.

12. Galen RS, G a m b i n o SR. Beyond N o r m a l i t y . T h e Pred ic t ive V a l u e a n d Eff ic iency of Medical Diagnosis . N e w York : J o h n Wiley a n d Sons, 1975 .

13. C l o u g h j D , El razak M, Ca labrese L H , Va lenzue l a R, B r a u n W E . W e i g h t e d c r i t e r i a a n d test select ion in t h e d iagnosis of SLF. (abst). Clin Res 1982; 3 0 : 8 0 4 A .

14. T a n EM, S c h u r P H , C a r r RI , Kunke l H G . Deoxyr ibonuc le i c acid ( D N A ) a n d a n t i b o d i e s t o D N A in s e r u m of pa t i en t s with systemic lupus e r y t h e m a t o s u s . J Clin Invest 1966; 4 5 : 1 7 3 2 -1740 .

15. C l o u g h j D , Va lenzue l a R. Re la t ionsh ip of rena l h i s topa tho logy in S L E neph r i t i s t o i m m u n o g l o b u l i n class of a n t i - D N A . A m J M e d 1980; 6 8 : 8 0 - 8 5 .



Documents

Serological tests for diagnosis of systemic lupus