Annals of the Rheumatic Diseases 1993; 52: 851-856

Predictive value of IgG autoantibodies againstC 1 q for nephritis in systemic lupus erythematosus

Carl E H Siegert, Mohammed R Daha, Carroll M E S Tseng, Ida E M Coremans,Leendert A van Es, Ferdinand C Breedveld

University Hospital,Leiden, TheNetherlandsDepartment ofRheumatologyC E H SiegertCMES TsengI E M CoremansF C Breedveld

Department ofNephrologyM R DahaL A van EsCorrespondence to:C E H Siegert,Department ofRheumatology, UniversityHospital, Building I C4R,PO Box 9600, 2300 RCLeiden, The Netherlands.

Accepted for publication24 August 1993

AbstractObjectives-Antibodies against Clq(CIqAb) have been demonstrated in theserum of patients with several immunecomplex diseases. Patients, particularlythose with lupus nephritis, were found tohave increased serum titres ofIgG ClqAbin a cross-sectional analysis. In thepresent prospective study correlationswere sought between serum titres of IgGClqAb and clinical as well as laboratoryparameters of disease activity in patientswith systemic lupus erythematosus(SLE).Methods-Titres of IgG ClqAb in theserum of 68 SLE patients were measuredserially during a three year period. At thesame time clinical and laboratory para-meters ofdisease activity were assessed.Results-Increased titres of IgG ClqAbwere found in the serum of 56% ofSLE patients during the study. Significantcorrelations were found betweenincreased titres of IgG ClqAb and renalinvolvement. Clinical signs of renalinvolvement were found to be associatedwith significant increases of serum titresofIgG ClqAb in the six months precedingthis appearance. Fifty per cent of theincreases in serum titres of IgG ClqAbwere followed by the development ofrenalinvolvement. Elevated serum titres ofIgG ClqAb were especially related toproliferative forms of glomerulonephritis.Furthermore, significant correlationswere found between serum titres of IgGClqAb and serum levels of immunecomplexes, levels of complement compo-nents, and titres ofantibodies to DNA.Conclusions-The results suggest that IgGClqAb play a pathogenic role in thedevelopment of lupus nephritis and thatserial measurement ofserum titres ofIgGClqAb is useful in the management ofSLE patients.

(Ann Rheum Dis 1993; 52: 851-856)

Autoantibodies to C I q (C I qAb) have beenreported in diseases in which immunecomplexes are considered to play a pathogenicrole, such as systemic lupus erythematosus(SLE), rheumatoid vasculitis, and membrano-proliferative glomerulonephritis.'-5 C IqAb ofboth immunoglobulin classes G and A have

been described.'2 4 The results of a cross-sectional study suggested that serum titres ofIgG Cl qAb are associated with renalinvolvement in SLE patients.6 The aim of thepresent study was to further investigateassociations between serum titres of IgGClqAb in SLE patients and clinical as well aslaboratory parameters of disease activity in aprospective study.

Materials and methodsPATIENTSSixty eight patients visiting the outpatientclinic of the department of rheumatology werestudied prospectively between April 1988 andMay 1991. All patients met the criteria for theclassification of SLE.7 The mean periodbetween the moment of diagnosis and themoment of study entrance was 6 years (range:1 month to 24 years). There were 65 femaleand three male patients. The mean age of thepatients at the start of the study was 38 years(range: 14-75 years). Clinical parameters ofdisease activity were recorded every six monthsby one observer (FCB). At the same time bloodsamples were obtained for the determination oflaboratory parameters of disease activity.Serum samples for the measurement of titresof IgG C lqAb and of antibodies againstdouble-stranded DNA (DNA-Ab) titres werestored in aliquots at -20°C until use. Fourpatients could not be evaluated for thecomplete study period.

CLINICAL PARAMETERSClinical parameters of disease activity werescored according to organ involvement.Erythema, photosensitivity, oral ulcers, discoidlesions, vasculitis, or alopecia were recordedfor the assessment of skin involvement.Arthritis was recorded for the assessment ofjoint involvement, and pericarditis, pleuritis, orperitonitis for the assessment of serosainvolvement. Proteinuria (+++ or >0 5 gr/24h), haematuria (> 10 red blood cells/high powerfield), or a decreased creatinine clearance(< 100 ml/min) were recorded as manifes-tations of renal involvement. Seizures, cere-brovascular accidents, myeloneuropathy, orsevere personality disorders were recorded forthe assessment of central nervous system(CNS) involvement. A cumulative score forgeneral disease activity (SLE-DAI) was used aswell.8 During the study non-steroidal anti-

851

on Decem

ber 5, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.52.12.851 on 1 D

ecember 1993. D

ownloaded from

Siegert, Daha, Tseng, Coremans, van Es, Breedveld

inflammatory drugs, corticosteroids, hydroxy-chloroquine, and cytostatic drugs were takenby 29%/ 76%, 40%, and 46% of the patients,respectively.

LABORATORY PARAMETERS

Levels of complement components Clq, C4,and C3 as well as total functional complementactivity (CH50) were measured according todescribed methods.9 Levels of circulatingimmune complexes were measured using thefluid phase C lq binding assay (Cl qBA) andexpressed a jig/ml of aggregated IgG.'0 "

Levels of DNA-Ab were measured using aCrithidia lucilae assay and expressed in titres oftwo-fold dilutions, starting at a dilution of1/10.12 A significant increase of serum titres ofDNA-Ab was defined as at least a doubling ofthe titre (initial titre at least 1/10). Titres ofDNA-Ab were also measured with the Farrassay in the serum of patients who developedrenal involvement. In this assay bound DNA-Ab are precipitated by ammonium sulphateand the results were expressed in internationalunits per millilitre serum (IU/ml) (DiagnosticProducts Los Angelos, CA).'3 14 The upperlimit of normal values in this assay was 15 IU/ml and a significant increase in DNA-Ab titreswas defined as an increase of at least 25% andat least 15 IU/ml.

C 1QAB MEASUREMENT

Titres ofIgG C 1qAb were measured by ELISAunder 1M NaCl conditions. In this assay onlyIgG antibodies to the collagen-like region ofC 1q are detected. Neither immune complexesnor rheumatoid factors influence the assayresults.4 Titres are expressed as U/ml related toa standard serum. The mean IgG ClqAb titre+ 2 standard deviations of serum samples of 80healthy blood donors was regarded as theupper limit of normal. This value was 90 U/ml.A significant increase of IgG C iqAb titres wasdefined as an increase of at least 25% and atleast 90 U/ml. Total IgG levels were measuredby ELISA in which wells were coated withmouse monoclonal anti-human IgG andsubsequently bound polyclonal IgG in serumdilutions was detected by horse-radishperoxidase conjugated rabbit anti-human IgG.

RENAL BIOPSIES

In 20 patients a biopsy had already beenobtained at a mean time of three years beforethe start of the study (range: two months to 14

Table 1 Clinical parameters of disease activity according to organ involvement of 68patients with SLE at study entrance and during three years offollow up

Organ Number of Number of episodes Number ofpatients (%)involvement patients (%o) with with new organ with episodes with new

organ involvement involvement during organ involvementat study entrance the study during the study

Skin 25 (37) 30 26 (38)Joints 5 (7) 10 8 (12)Serosa 5 (7) 7 7(10)Central nerve system 11 (16) 13 11 (16)Kidney 21(31) 7 7(10)Proteinuria 18 (26) 6 6 (9)Haematuria 16 (24) 6 6 (9)

years). In five patients a biopsy was obtainedduring the study. Patients had a renal biopsywhen clinically indicated. The World HealthOrganisation classification of lupus nephritiswas used to describe the histological lesions.15Mixed forms of glomerulopathy were classifiedaccording to the proliferative lesion present.

STATISTICSThe Chi-square, Fisher's exact test (fornumbers <5), and the Spearman non-parametric rank correlation coefficient wereused. P-values below 0 05 were considered tobe significant. Since the same number ofobservations was made for each patient at fixedintervals, each observation was considered as aseparate observation when the Spearmancorrelation coefficient was used. A total of 422observations were made. Clinical parameterswere scored as qualitative variables. Lab-oratory parameters were considered asquantitative variables. C lqAb titres wereconsidered both as qualitative (normal andincreased titres) and as quantitative variables.

ResultsCLINICAL PARAMETERSThe clinical parameters of disease activityobserved in 68 SLE patients at the start of thestudy as well as new disease manifestationsobserved during three years of follow up arepresented in table 1. Haematuria andproteinuria, as manifestations of renalinvolvement, are also shown. Skin and renalinvolvement were disease manifestations withthe highest prevalence at the start of the study.New clinical manifestations of organinvolvement during follow up predominantlyconcerned the skin. Joint, serosa, CNS, andrenal involvement occurred less frequently.New involvement of skin, joints or CNSoccurred twice in 4, 2, and 2 patientsrespectively.

Increased serum titres of IgG ClqAb werefound in the serum of 29 (43%) patients at thestart of the study and in the serum of 38 (56%)patients at any time during the study.Comparison of increased serum titres of IgGC lqAb and clinical parameters of diseaseactivity of SLE at study entrance demonstrateda significant association between titres of IgGCl qAb and renal involvement (table 2).Similar associations were found between thepresence of increased serum titres of IgGC 1qAb and haematuria as well as proteinuria.No association was found between thepresence of increased titres of IgG C 1qAb andthe involvement of other organs at the start ofthe study. Increased titres of IgG C lqAbmeasured at fixed intervals during three yearsof follow up were significantly correlated withthe SLE-DAI score for general disease activitywhich were measured at the same time(r = 0'18; p < 0-01). Comparison of increasedserum titres of IgG ClqAb and the presenceof organ involvement during three years offollow up only revealed significant associationsbetween the presence of increased IgG CIqAb

852

on Decem

ber 5, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.52.12.851 on 1 D

ecember 1993. D

ownloaded from

Predictive value of IgG autoantibodies against Clq for nephritis in systemic lupus erythematosus

Table 2 Comparison of the presence of increased serumtitres ofIgG ClqAb with the presence oforgan involvementat study entrance in 68 patients with SLE

Increased ClqAb titre + + - p value*Presence of + +organ involvement

Skin 7 23 20 18 0111Joints 2 36 27 3 0-64Serosa 4 38 25 1 010Central nerve system 3 31 26 8 0 22Kidney 15 33 14 6 <001Proteinuria 14 35 15 4 <0 001Haematuria 12 35 17 4 <0 01

*Chi-square or Fisher's exact test.

titres and the occurrence of renal involvement(r = 0 40; p < 0 00 1). No association wasfound between serum titres of IgG C I qAb andthe use of particular drugs for the treatment ofSLE (data not shown).

LABORATORY PARAMETERS

Comparison between serum titres of IgGC 1qAb and laboratory parameters of SLEactivity demonstrated positive correlationsbetween IgG C iqAb titres and serum levels ofcreatinine, levels of circulating immunecomplexes measured by the fluid phaseC 1qBA, and serum titres ofDNA-Ab (table 3).Negative correlations were found betweenserum titres of IgG C 1qAb and levels ofcomplement components of the classicpathway as well as total functional complementactivity (CH50). No correlation was foundbetween serum titres of IgG Cl qAb and theESR, haemoglobin content, leucocyte count,platelet count, and total IgG levels.

RELATION BETWEEN THE DEVELOPMENT OF

ORGAN INVOLVEMENT AND INCREASES IN

SERUM TITRES OF IGG C 1QABTo investigate a possible pathogenetic role ofIgG Cl qAb in the development of diseasemanifestations ofSLE, the association betweenthe appearance of clinical signs of organinvolvement and significant increases in serumtitres of IgG Cl qAb was explored during thesix months preceding such appearances (table4). A significant increase in titres of IgGCiqAb was observed 10 times in 10 patients.Increases in serum titres of IgG C 1qAb weresignificantly associated with the developmentof renal involvement. Each patient who

Table 3 Correlation * between serum titres ofIgG ClqAband laboratory parameters ofdisease activity in 68 patientswith SLE during three years offollow up

Laboratory parameter Correlation p valuecoefficient

ESR 0-06 0-18Haemoglobulin level -008 0 10Leucocyte count -002 0-61Platelet count 0.01 0 74Serum creatinine 0-12 <0 05Clq -019 0 001C4 -017 <0 01C3 -015 <0 01CH50 -015 <0.01Fluid phase ClqBA 0 21 <0.01DNA-Ab 0-14 <0.01

*Spearman rank correlation coefficient.

Table 4 Comparison of increases in serum titres ofIgGClqAb with the development oforgan involvement in thesix months following this increase in 68 patients with SLEduring three years offollow up

Increased ClqAb titre + - + - p value*Development of + +organ involvement

Skin 4 36 6 26 0.59Joints 1 51 9 9 0 56Serosa 3 54 7 4 0 06Central nerve system 0 47 10 13 0 10Kidney 5 56 5 2 <0 001Proteinuria 4 56 6 2 <0 01Haematuria 5 57 5 1 <0 001

*Fisher's exact test.

developed renal involvement associated withthe presence of increased serum IgG ClqAbtitres had a significant titre increase precedingthe first appearance of clinical manifestation.Two patients who did not have increased titresof IgG Cl qAb developed renal involvement.No correlation was found between increases ofserum titres of IgG C IqAb and new SLEmanifestations in other organs.The specificity, sensitivity, and predictive

value of significant increases of serum titres ofIgG ClqAb for the development of kidneyinvolvement are present in table 5. Significantincreases in titres of IgG C 1qAb were followedby the development of kidney involvement in50% of the cases.

BIOPSIES

Renal biopsies were obtained in 20 SLEpatients before the start of the study and in fivepatients during the study (table 6). In mostbiopsies a focal segmental (Class III) or diffuseproliferative (Class IV) glomerulonephritis wasfound. In the serum of 72% of patients whohad glomerular lesions classified as Class III orIV before the study, increased serum titres ofIgG Cl qAb were found at the start of thestudy. In two other patients with glomerularlesions classified as Class I and II, titres of IgGClqAb at the start of the study were found tobe within the normal range. Five patients hada renal biopsy during the study. Three biopsieswere classified as Class IV and one biopsy wasclassified as Class III. The serum of these fourpatients contained increased titres of IgGCl qAb before and at the time of the biopsy.The biopsy of the other patient, who did nothave increased titres ofIgG C IqAb throughoutthe study period, was classified as Class IIA.

RENAL INVOLVEMENT AND TITRES OF DNA-ABPrevious studies have reported a relationbetween serum DNA-Ab and renalinvolvement in SLE.14 1618 We thereforefocussed on levels of DNA-Ab as measured bythe C luciliae assay and the Farr assay of thepatients with renal involvement. Comparisonof increased serum levels of DNA-Ab asmeasured by the C luciliae assay with renalinvolvement measured at all observation pointsthroughout the study demonstrated asignificant correlation (r = 0-24; p < 0 0001).Three of the seven SLE patients who

853

on Decem

ber 5, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.52.12.851 on 1 D

ecember 1993. D

ownloaded from

Siegert, Daha, Tseng, Coremans, van Es, Breedveld

Table 5 The sensitivity, specificity, andpredictive value (95% confidence interval) ofasignificant increase in serum titres ofIgG ClqAbfor the development of renal involvementor its manifestations proteinuria and haematuria in 68 patients with SLE during threeyears offollow up

Sensitivity Specificity Positive Negativepredictive value predictive value

Renal involvement 71% (29-96) 92% (82-97) 50% (19-81) 97% (88-99)Proteinuria 67 (22-96) 90 (80-96) 40 (12-74) 57 (88-99)Haematuria 83 (36-99) 92 (82-97) 50 (19-81) 98 (91 -99)

Table 6 Renal biopsies in 25 patients with SLE classified according to the WHO andrelated to the presence of increased serum titres ofIgG ClqAb during the study

WHO Biopsies Increased ClqAb Biopsies Increased ClqAbclass before titre at study during titre at the time

follow up entrance follow up of biopsy

I 1 0 0 0IIA 1 0 1 0IIB 0 0 0 0III 4 2 1 1IV 14 11 3 3V 0 0 0 0

developed renal involvement during the studyhad increased serum DNA-Ab levels.However, significant increases in serum levelsDNA-Ab before the development of renalinvolvement were observed less frequentlycompared with rises of serum IgG C lqAbtitres. When employing the C luciliae assay, a

significant titre increase was observed in onlyone of the patients and with the Farr assay sucha rise was observed in the same and in twoother patients. In all these three patients a

significant rise of Cl qAb titres was alsoobserved.

DiscussionThe existence of monomeric IgG that binds toC 1 q had already been noted a long time beforethe existence of IgG Cl qAb was established.When studying the nature of immune

complexes in SLE patients, several studiesreported that not only immune complexes butalso small '7S materials' with characteristicssimilar to monomeric IgG bound to C q.'9-23It was also noted that high titres of C 1qbinding '7S materials' were frequently presentin the serum ofSLE patients and their presencewas related to proliferative glomerularlesions.2' 23

The aim of our prospective study was tofurther investigate associations between serum

titres ofIgG C 1qAb and disease manifestationsof SLE. Significant associations were foundbetween increased serum titres of IgG Cl qAband nephritis, with subsequent loss of kidneyfunction. In addition, the development ofnephritis was frequently preceded by a

significant increase in serum titres of IgGCl qAb.Many studies have investigated associations

between disease activity of SLE and laboratoryparameters. Levels of separate complementcomponents,'6 17 24 25 circulating immunecomplexes,'7 24 26 27 and DNA-Ab14 16-18 were

reported to be related to disease activity, butthe results were not uniform. These laboratoryparameters were most strongly related to lupusnephritis.'6 18 26 27 The present study suggeststhat IgG C 1qAb might be seen as a new

parameter for the development of lupusnephritis. As the appearance of renalmanifestations associated with IgG C 1qAb wasalways preceded by a significant titre increaseand since the predictive value of increases ofserum titres of IgG C IqAb for thedevelopment of kidney involvement wasrelatively high, the serial measurement ofserum titres of IgG Cl qAb might be useful inthe management of SLE patients.Measurement of DNA-Ab titres is used for

the classification of SLE and in many centresas a laboratory parameter of disease activity.Several assays are available, each of whichdetects a part of the total spectrum of DNA-Ab.28 The immunofluorescent assay using Cluciliae, employed throughout this study, andthe Farr assay are the most commonly usedassays. Two recent prospective studiesdemonstrated that increments of DNA-Abtitres were associated with general diseaseactivity and with the development ofnephritis.'4 18 Comparison between the Cluciliae assay and the Farr assay demonstratedthat the Farr assay was the most sensitive assayin this respect.'4 The development of kidneyinvolvement was preceded by increases in titresof DNA-Ab in nine of the 13 patientsmeasured by the C luciliae assay and in 12patients measured by the Farr assay. In theprevious two studies serum samples for themeasurement ofDNA-Ab titres were obtainedmore frequently than in the present study,which, in addition to differences in the patientpopulation, may be an explanation for thecontrasting results.C 1qAb titres were correlated with several

laboratory parameters known for theirassociation with disease activity of SLE. Thecorrelation between serum titres of IgG C 1qAband hypocomplementaemia as well as levels ofcirculating immune complexes may eitherindicate a contribution of IgG C I qAb toimmune complex formation and subsequentcomplement consumption, or a directinfluence on activation of the classic pathwayof the complement system by IgG Cl qAb. Ina previous study we could not demonstrate thatIgG C 1qAb had a direct influence onactivation of the classic pathway of thecomplement system in SLE patients.29However, it could be demonstrated thatcirculating immune complexes in the serum ofpatients with rheumatoid vasculitis containCl qAb.4 The established correlation thereforebetween titres of IgG CiqAb and levels ofcirculating immune complexes and comple-ment consumption indicates that IgG Cl qAbin the serum of SLE patients contributes toimmune complex formation.Although C I qAb titres were measured in the

majority of patients after renal biopsies wasobtained, the association between IgG CiqAband Class III and IV lupus nephritis may beimportant since the majority of the patho-logical diagnoses in lupus nephritis does notchange with time.30 The association foundbetween IgG C IqAb and the proliferative formof glomerulonephritis in SLE patients studiedboth retrospectively and prospectively adds to

854

on Decem

ber 5, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.52.12.851 on 1 D

ecember 1993. D

ownloaded from

Predictive value of IgG autoantibodies against Clq for nephritis in systemic lupus erythematosus

similar observations in lupus and other formsof nephritis. 1 5 21 23 31 These observationssuggest a pathogenic role for IgG C l qAb in thedevelopment of glomerulonephritis. Immunecomplexes are generally considered to play a

pathogenetic role in lupus nephritis.32Microscopic studies demonstrated thepresence of immunoglobulins, Clq, and othercomplement components in glomeruli of SLEpatients who develop nephritis.33 Deposition ofimmune complexes in the kidney is believed toresult from either glomerular trapping ofcirculating immune complexes or from in situformation of immune complexes by fixation ofantibodies to endogenous or exogenous

(planted) antigens on the glomerular basementmembrane. The degree to which eachmechanism is responsible for the developmentof lupus nephritis is uncertain. Circulatingimmune complexes are found in high titres inthe serum of nearly all SLE patients withproliferative glomerulonephritis.2' 23 Theseimmune complexes were shown to containClq.34 A possible way of in situ immunecomplex formation is preferential binding ofimmune complexes to the glomerularbasement membrane on the basis of charge-charge interactions.35 Cationic immunecomplexes were shown to deposit principally atanionic sites of the glomerular basementmembrane.36 37 Cationic antigens may alsodeposit at anionic sites of the glomerularbasement membrane before forming an

immune complex.38 39 C 1 q is a cationic proteinand may induce local immune complexformation in the presence of C 1qAb once it isbound to a solid-phase in the kidney. Recentstudies have demonstrated that C lqAb are

able to bind to C lq that is present on theglomerular basement membrane in vivo.40 41

IgG C l qAb may thus play a role in thedevelopment of nephritis in SLE patients,either by a contribution to the formation ofcirculating immune complexes that deposit inthe kidney or by local formation of immunecomplexes on the glomerular basementmembrane.

In conclusion the present prospective studydemonstrates an association between thepresence of increased serum titres of IgGCl qAb and the development of proliferativeglomerulonephritis in patients with SLE. Theassociation in time between increases in titreand the development of kidney involvementsuggests that serial measurment of serum titresof IgG C lqAb may prove to be a valuable toolin the management of SLE patients.

The authors greatly appreciate the contribution of Dr JHermans, Department of Medical Statistics, UniversityHospital, Leiden, to the analysis of the data.

1 Uwatoko S, Aotsuka S, Okawa M, Egusa Y, et al.Characterization of C lq-binding IgG complexes insystemic lupus erythematosus. Clin ImmunolImmunopathol 1984; 30: 104-16.

2 Uwatoko 5, Aotsuka S, Okawa M, et al. C lq solid-phaseradioimmunoassay: evidence for detection of antibodydirected against the collagen-like region of C lq in serafrom patients with systemic lupus erythematosus. ClinExp Immunol 1987; 69: 98-106.

3 Antes U, Heinz H-P, Loos M. Evidence for the presence ofautoantibodies to the collagen-like region of C q insystemic lupus erythematosus. Arthritis Rheum 1988; 31:457-64.

4 Siegert C E H, Daha M R, Voort E A M van der, BreedveldF C. IgG and IgA antibodies to the collagen-like regionofC Iq in rheumatoid vasculitis. Arthritis Rheum 1990; 33:1646-54.

5 Strife C F, Leahy A E, West C D. Antibody to a cryptic,solid-phase C lq antigen in membranoproliferativenephritis. Kidney Int 1989; 35: 836-42.

6 Siegert C E H, Daha M R, Westedt M L, Voort E A M vander, Breedveld F C. IgG autoantibodies against Cl q arecorrelated with nephritis, hypocomplementemia, anddsDNA antibodies in systemic lupus erythematosus. JRheumatol 1991; 18: 230-4.

7 Tan E M, Cohen A S, Fries J F, et al. The 1982 revisedcriteria for the classification of systemic lupuserythematosus. Arthritis Rheum 1982; 25: 1271-7.

8 Chang C-H, Urowitz M B, Gladman D, Bombardier C.Impact of SLE disease activity on survival with validationof a disease activity index. Arthritis Rheumn 1988; 31: S56.

9 Daha M R, Berdina R M, Thomson J, et al. Combinedhereditary deficiency of the sixth component ofcomplement and factor VIII coagulant activity in a Dutchfamily. Clin Exp Immunol 1982; 48: 733-8.

10 Zubler R H, Lange G, Lambert P H, Miescher P A.Detection of immune complexes in unheated sera bymodified 1251-binding test. J Immunol 1976; 11: 232-5.

11 Kauffman R H, Es L A van, Daha M R. Aggregated humanimmunoglobulin G stabilized by albumin: a standard forimmune complex detection. J Immunol Methods 1979; 31:11-22.

12 Aarden L A, Groot E R de, Feltkamp T E W. Immunologyof DNA III: Crithidia Luciliae, a simple substrate for thedetermination of anti-ds-DNA with the immuno-fluorescent techinque. Proc N Y Acad Sci 1975; 254:505-15.

13 Feltkamp T E W, Kirwood T B L, Miani R N, Aarden LA. The first international standard for antibodies todouble stranded DNA. Ann Rheumn Dis 1988; 47: 740-6.

14 Borg E J ter, Horst G, Hummel E J, Limburg P D,Kallenberg C G M. Measurement of increases in anti-double-stranded DNA antibody levels as predictor ofdisease exacerbations in systemic lupus erythematosus.Arthritis Rheum 1990; 33: 634-43.

15 McCluskey R T. Lupus nephritis. Sommer S C, eds. Kidneypathology decennial 1966-1977. New York: Appleton-Century-Crofts, 1975: 435.

16 Schur P H, Sandson J. Immunologic factors and clinicalactivity in systemic lupus erythematosus. N Eng J Med1968;278: 533-8.

17 Lloyd W, Schur P H. Immune complexes, complement, andanti-DNA in exacerbations of systemic lupuserythematosus (SLE). Medicine 1981; 60: 208-17.

18 Swaak A J G, Groenwold J, Bronsveld W. Predictive valueof complement profiles and anti-DNA in systemic lupuserythematosus. Ann Rheum Dis 1986; 45: 359-66.

19 Agnello V, Koffler D, Eisenberg J W, Winchester R J, KunkelH G. C I q precipitins in the sera of patients with systemiclupus erythematosus and other hypocomplementemicstates: characterization of high and low molecular weighttypes. J Exp Med 1971: 134: 228S-41S.

20 Robinson M F, Roberts J L, Verrier Jones J, Lewis E J.Circulating immune complex assays in patients with lupusand membranous glomerulonephritis. Clin ImmunolImmunopathol 1979; 14: 348-60.

21 Tung K S K, DeHoratius R J, Williams R C. Study ofcirculating immune complex size in systemic lupuserythematosus. Clin Exp Immunol 1981; 43: 615-25.

22 Hack C E, Belmer A J M. The IgG detected in the Cl qsolid-phase immun-complex assay is not always ofimmune complex nature. Clin Immunol Immunopathol1986; 38: 120-8.

23 Wener M H, Mannik M, Schwartz M M, Lewis E J.Relationship between renal pathology and the size ofcirculating immune complexes in patients with systemiclupus erythematosus. Medicine 1987; 66: 85-97.

24 Valentijn R M, Overhagen H van, Hazevoet H H, et al. Thevalue ofcomplement and immune complex determinationin monitoring disease activity in patients with systemiclupus erythematosus. Arthritis Rheum 1985; 28: 904-13.

25 Cameron J S, LessofM H, Ogg C S, Williams B P, WilliamsD G. Disease activity in the nephritis of systemic lupuserythematosus in relation to serum complementconcentrations. Clin Exp Immunol 1976; 25: 418-27.

26 Levinsky R J, Cameron J S, Soothill J F. Serum immunecomplexes and disease activity in lupus nephritis. Lanceti: 564-7.

27 Cairns S A, London A, Mallick N P. The value of threeimmune complex assays in the management of systemiclupus erythematosus: an assessment of immune complexlevels, size and immunochemical properties in relation todisease activity and manifestations. Clin Exp Immunol1980; 40: 273-82.

28 Smeenk R, Lelij G van der, Aarden L. Avidity of antibodiesto dsDNA: comparison of IFT on Crithidia luciliae, Farrassay, and PEG assay.J_Inmunol 1982; 128: 73-8.

29 Siegert C E H, Daha M R, Lobatto S, Voort E A M vander, Breedveld F C. IgG antibodies to C 1q do notdetectably influence activation of complement in vivo andin vitro patients with systemic lupus erythematosus.Immunol Res 1992; 11: 91-7.

30 Appel G B, Silva F G, Pirani C L, Meltzer J I, Estes D.Renal involvement in systemic lupus erythematosus(SLE): a study of 56 patients emphasizing histologicclassification. Medicine 1978; 57: 371-410.

31 Siegert C E H, Daha M R, Voort E A M, Halma C,Breedveld F C. IgG and IgA antibodies to C 1q in

855

on Decem

ber 5, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.52.12.851 on 1 D

ecember 1993. D

ownloaded from

Siegert, Daha, Tseng, Coremans, van Es, Breedveld

systemic and renal disorders. Clin Exp Rheumiiatol 1992;10: 19-23.

32 Mannik M. Pathophysiology of circulating immunecomplexes. Arthritis Rheurn 1982; 25: 783-7.

33 Agnello V, Koffler D, Kunkel H G. Immune complexsystems in the nephritis of systemic lupus erythematosus.Kidney Ilt 1973; 3: 90-9.

34 Greisman S G, Redechan P B, Kimberley R P, Christian CL. Differences among immune complexes: association ofClq in SLE immune complexes with renal disease.Imnmlunol 1987; 138: 739-45.

35 Couser W G, Salant D J. In situ immune complex formationand glomerular injury. Kidney Int 1980; 17: 1-13.

36 Caulfield J P, Farquhar M G. Distribution of anionic sitesin glomerular basement membranes: their possible role infiltration and attachment. Proc Natl Acad Sci USA 1976;73: 1646-50.

37 Gallo S R, Caulin-Glaser T, Lamm M E. Charge ofcirculating immune complexes as a factor in glomerular

basement localization in mice. Clin Invest 1981; 67:1305-13.

38 Oite T, Batsford S R, Mihatsa M Y, Takamiya H, Vogt A.Quantitative studies of in situ immune complexglomerulonephritis in the rat induced by planted,cationized antigen. _Exp Med 1982; 155: 460-74.

39 Vogt A, Rohrboch R, Shimizu F, Takamiya H, Batsford SR. Interaction of cationized antigen with rat glomerularbasement membrane: in situ immune complex formation.Kidney Itlt 1982- 22: 27-35.

40 Uwatoko S, Gauthier J, Mannik M. Autoantibodies to thecollagen-like region of Cl q deposit in glomeruli via Cl qin immune deposits. Clin Imm?77lunol Inimiunopathol 1991;61: 268-73.

41 Coremans I E M, Bruyn J A, Heer E de, Voort E A M van

der, Breedveld F C, Daha M R. Antibodies to Cl q causeglomerular injury in mice (abstr). Innnunobiol 1992; 184:417.

856

on Decem

ber 5, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.52.12.851 on 1 D

ecember 1993. D

ownloaded from