340 J Clin Pathol 2001;54:340–347

Leaders J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from

Autoantibodies and overlap syndromes in autoimmune rheumatic disease

E C Jury, D D’Cruz, W J W Morrow

Abstract . This article will consider, primarily, Many patients diagnosed with autoimmune overlap syndromes associated with autoim- rheumatic disease cannot be categorised mune rheumatic disease; specific reference will easily into one of the established clinical be made to profiles and ramifica- entities such as systemic erythema- tions for diagnosis and pathogenesis will be dis- tosus, , or systemic sclero- cussed. It should be noted that overlap between sis. The term “” has been specific autoimmune syndromes, such increasingly used to identify such patients as myasthenia gravis, Hashimoto’s thyroiditis, and is useful in terms of clarifying progno- and insulin dependent diabetes mellitus, is sis and facilitating disease management. frequently seen; however, a detailed discussion This article reviews overlap syndrome in of these conditions is beyond the scope of this autoimmune rheumatic disease, with paper. particular emphasis on the associated sero- Autoimmune rheumatic diseases are classified logical markers. using internationally accepted criteria, which (J Clin Pathol 2001;54:340–347) frequently incorporate the detection of specific Keywords: ; overlap syndromes; autoantibodies as unique diagnostic markers. autoimmune rheumatic disease Examples include: antidouble stranded DNA Centre for (anti-dsDNA) and anti-Smith antigen (anti- Research, University The complexity of is Sm) for establishing a diagnosis of systemic 1 College, London formidable and although the precise role of lupus erythematosus (SLE) ; the detection of W1T 4JF, UK autoantibodies in immunopathological proc- in the diagnosis of rheuma- E C Jury 2 esses remains uncertain, there is little doubt toid (RA) ; anticentromere, anti-DNA http://jcp.bmj.com/ topoisomerase I, and anti-RNA polymerase I Louise Coot Lupus that they contribute to disease advancement. Unit, St Thomas’s The formation of autoantibodies is a normal and III antibodies as markers for the sclero- Hospital, London physiological process: however, excessive pro- derma spectrum of disease; and a group of anti- SE1 7EH, UK duction of such antibodies can be harmful, bodies including antiliver/ microsomes D D’Cruz resulting in disease or injury. Autoantibodies (anti-LKM), antismooth muscle antigen (anti- may cause damage by reacting directly against a SMA), and antinuclear (ANA), included in Department of diagnostic criteria for autoimmune hepatitis.3 , St specific tissue or by the formation and deposi- on September 25, 2021 by guest. Protected copyright. Bartholomew’s and tion of circulating immune complexes in the However, many patients cannot be assigned The Royal London kidney, skin, , and nervous system. The to a single disease category. This diYculty has School of production of autoantibodies might be second- led to the concept of “overlap syndrome”, and ary to tissue damage or induced by independent where symptoms from two or more autoim- W J W Morrow pathogenic factors; organ specific and/or non- mune conditions are identified in the same Correspondence to: organ specific autoantigens may be involved. patient. Overlap syndromes are commonly Ms Jury Autoimmune rheumatic disorders are char- associated with autoimmune rheumatic dis- [email protected] acterised by autoantibodies to non-organ spe- ease, where up to 25% of patients with early Accepted for publication cific antigens, specifically, antigens occurring in stage or mild variant disease fall into this 20 September 2000 nucleated cells or among circulating plasma category.45 Overlap syndrome has also been described in patients with autoimmune liver Table 1 Overlap syndromes in autoimmune rheumatic disease and associated autoantibodies78 diseases, including autoimmune hepatitis, pri- mary biliary , and primary sclerosing 6 Overlap syndrome Autoantibody directed to cholangitis (table 1). Mixed connective tissue disease nRNP; hnRNP(A2/RA33 complex); U1 Clinically, it is useful to define overlap RNA syndromes to clarify prognosis and facilitate dis- tRNA synthetase syndrome Jo-1, PL-7, PL-12, OJ, EJ ease management.9 Two approaches can be /pulmonary fibrosis KJ Polymyositis/systemic lupus erythematosus 56 kDa nRNP made when categorising such conditions. The Polymyositis/ Ku; PM-Scl; U2 RNP; DNA-PK first is by the identification of a pattern of clini- Secondary Sjögren’s syndrome Ro/SSA; La/SSB; Ki/SL cal features, and a good example is the tRNA Rheumatoid arthritis/systemic lupus erythematosus None identified Scleroderma (CREST)/primary biliary cirrhosis Centromere; pyruvate dehydrogenase synthetase syndrome. The second is by the complex initial detection of a unique autoantibody profile Systemic lupus erythematosus/autoimmune hepatitis tRNA combined with specific clinical findings—for hnRNP, heterogeneous nuclear ribonuclear ; PM-Scl, polymyositis-scleroderma overlap example, mixed connective tissue disease syndrome. (MCTD).10

www.jclinpath.com Autoantibodies and overlap syndromes in autoimmune rheumatic disease 341

The most abundant snRNPs are denoted U1, Autoimmune rheumatic disease U2, U4/U6, and U5 and, together with J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from Connective tissue diseases (CTDs) are hnRNP, they form a major component of the characterised by considerable clinical diversity spliceosome, where premessenger RNA is and heterogeneity. Characteristic clinical fea- processed into mature mRNA. The snRNP tures and the detection of autoantibodies help spliceosomal complex has several associated to define these disorders and facilitate diagno- common proteins that might also be autoanti- sis and appropriate treatment. The CTD over- genic. Common proteins binding to U1 snRNP lap syndromes are distinguished by the con- particles are represented as U1–70K, U1A, and comitant occurrence of clinical and serological U1C. Antibodies can be directed to several features of the component diseases. Any CTD components of the spliceosome, although with can be a partner in an overlap disorder—for diVerent specificities in diVerent diseases.21 29 example, RA is a common partner to SLE, sys- Antibody specificity is normally determined by 11 12 temic sclerosis, or Sjögren’s syndrome (SS). western blot analysis. Autoantibodies binding In some cases, overlap syndromes have not to spliceosomal snRNPs (U1, U2, U4/U6, and been clearly circumscribed or widely accepted U5 complexes) are termed anti-Sm antibodies and the term “undiVerentiated” or “transi- and are associated with SLE. Antibodies tional” CTD has been used, implying that they recognising the U1 snRNP complex alone are are incompletely developed “classic” CTD, termed anti-nRNP and are considered the rather than distinct entities with overlapping hallmark of MCTD (reviewed by Klein Gun- 4 features. Disease in patients with undiVerenti- neweik and colleagues30). Anti-nRNP antibod- ated CTD might evolve into a more recognis- ies recognise primarily the U1 snRNP specific able syndrome, most commonly SLE, , proteins U1–70K, U1A, and U1C and U1 13–15 or scleroderma, although one report sug- snRNA. There are conflicting reports detailing 16 17 gested that most remain undiVerentiated. the role of anti-nRNP antibodies in predicting Overall, the picture of overlap syndromes with disease severity and progression. On balance, respect to CTD is complex and heterogeneous. evidence suggesting that U1 snRNP antibody Observer bias might play a role in disease classi- titres correlate with clinical events is uncon- fication, so the presence of specific autoantibody vincing.31 Alternatively, antibodies to U1 RNA profiles might be a useful tool in the diagnosis have been shown to be associated with and management of such patients. progressive disease and exacerbation.27 Epitope spreading, whereby an immune response to an Mixed connective tissue disease epitope expands to include associated epitopes The existence of MCTD has been the subject on the same molecule, is implicated in the of much debate10 15 18–20 (reviewed by Smolen immune response to snRNA and the produc- and Steiner21). A detailed discussion is beyond tion of possibly pathogenic antibodies associ- the scope of this review, but suYce to say this ated with disease. Recent reports suggest that

syndrome was originally described in 1972 by secondary light chain rearrangements in U1- http://jcp.bmj.com/ Sharp and colleagues.22 The syndrome is a RNA reactive B cells, possibly as a result of combination of features typically found in crossreactivity with viral immunogens, can lead patients with SLE, scleroderma, polymyositis/ to changing specificity and aYnity of anti-RNA dermatomyositis (PM/DM), or RA. The pres- antibodies (epitope spreading).32 It is interest- ence of high titre ANA with speckled pattern ing to note that long term follow up of clinical and antibodies to uridine rich U1 small nuclear and serological findings in patients with

ribonucleoprotein (snRNP) diVerentiates this MCTD has demonstrated intramolecular on September 25, 2021 by guest. Protected copyright. MCTD from other CTDs. Three points spreading of autoantibody reactivity against provide a focus for controversy: first, patients snRNP polypeptides. This was followed by with MCTD may ultimately develop an estab- “epitope contraction” and the ultimate disap- lished CTD, commonly, systemic sclerosis, pearance of anti-snRNP antibodies during SLE, or myositis.13–15 Second, anti-U1 snRNP prolonged remission in some patients.33 antibodies are not confined to MCTD and are Spliceosomal hnRNP, a collection of 30 pro- found in patients with other CTDs, especially teins termed hnRNP A1 to U, is also a SLE,19 23 and, finally, the definition of MCTD common target for autoantibodies in several is confusing. Although originally described as a CTDs, namely, SLE, MCTD, and RA.30 34 benign disorder, considerable organ involve- Autoantibodies to hnRNP A/B (RA33 com- ment can accompany this condition and the plex)26 are found in sera from patients with RA, prognosis may be poor.10 SLE, and MCTD (table 2). In each disease, Several classification criteria exist for diverse hnRNP complex epitopes are MCTD22 24 25: all regard the presence of high titre speckled ANA and anti-U1 snRNP as Table 2 hnRNP,Sm, and U1 snRNP reactivities of essential for diagnosis. However, other auto- autoantibodies in patients with SLE, RA, and MCTD antibodies are seen, specifically antibodies to (adapted from Smolen and Steiner21) heterogeneous nuclear ribonucleoprotein (hnRNP),26 U1 RNA,27 phospholipid,28 rheu- Autoantibody MCTD RA SLE matoid factor, and endothelial cells.29 Antibod- Anti-hnRNP + + + Anti-Sm − − + ies to dsDNA and Sm, both specific markers Anti-U1 snRNP + − + for SLE, are rare.21 It is well established that snRNPs are a group hn, heterogeneous nuclear; MCTD, mixed tissue connective disease; RA, rheumatoid arthritis; RNP, ribonuclear protein; of protein particles associated with small RNA SLE, systemic lupus erythematosus; Sm, Smith antigen; sn, molecules in the nucleus of all eukaryotic cells. small nuclear.

www.jclinpath.com 342 Jury, D’Cruz, Morrow

recognised—for example, in MCTD, sera aminoacyl-tRNA synthetases have been de- appear to recognise a distinctive private epitope scribed: anti-Jo-1 (histidyl), PL-7 (threonyl), J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from on the hnRNP A2 protein.35 PL12 (alanyl), OJ (isoleucyl), and EJ (glycyl). It has been mentioned previously that a Each antibody is independently associated with diagnostic feature of MCTD is the production a distinctive clinical syndrome exhibiting simi- of large quantities of autoantibodies. An imbal- lar clinical features; the synthetase antibodies ance in synthesis could be an share immunochemical properties including important factor in driving the immune inhibition of enzymatic function; they do not response, resulting in excessive autoantibody crossreact with other synthetases; only one production. Hassan and colleagues36 have antisynthetase antibody will occur in any one reported an increase in both type 1 and type 2 patient and they have strong immunogenetic in patients with MCTD, including an associations. The most predominant antibodies increase in the cytokines interleukin 10, are directed towards Jo-1 (IgG1 subclass); interferon ã, and tumour factor á. patients with these antibodies tend to have a Immunogenetic studies have established a relatively young age of disease onset compared linkage between antigens of the major histo- with other myositis conditions, arthritis is a compatibility complex (MHC), human leuco- more common feature, and these antibodies cyte antigens (HLA) DR4 and DR2, and the are inclined to be associated with PM rather presence of autoantibodies to U1 snRNP and than DM.40 42 U1 snRNA, with the core clinical features of Mechanisms of autoimmunity are focused MCTD.18 37 In the future, genetic studies might around the concept that MSA play a funda- be able to predict those patients with overlap mental role in the disease process. Antisyn- syndrome who will eventually develop a classi- thetase antibodies each inhibit the enzyme of cally defined CTD such as scleroderma, SLE, their antigen specificity, indicating the highly or myositis. selective nature of the autoimmune response. Enzyme inhibition is proportional to the Polymyositis, dermatomyositis, and amount of antibody detected in patients’ sera overlap syndromes and there are some reports to suggest that anti- Autoimmune myositis has two major classifica- body titre correlates with disease activity. The tions, polymyositis (PM) and dermatomyositis autoimmune response is antigen driven, possi- (DM). Both cause weakness of the proximal bly initiated by antigen mimicry, with infec- muscles and patients have a high frequency of tious agents mimicking host antigens and specifically associated autoantibodies to nu- inducing an immune response capable of clear and cytoplasmic antigens, termed myosi- crossreaction with host proteins (molecular tis specific antibodies (MSA) (reviewed by mimicry). Various infectious agents have been Mimori38). Myositis specific antibodies are of implicated including enteroviruses, such as fundamental importance: they are a tool for Coxsackie virus, picornaviruses, and retro-

diagnosis and patient classification and provide viruses. Host and viral protein or RNA could http://jcp.bmj.com/ an insight into the pathogenesis of these condi- interact forming altered host proteins that are tions (for reviews see TargoV39 40). MSA are immunogenic; picornaviruses have been impli- found almost exclusively in patients with cated in this mechanism. Alternatively, Plotz47 DM/PM and associated overlap syndromes. suggested that anti-idiotype mechanisms might Autoantibodies detected in myositis associated drive the autoimmune response in myositis overlap syndromes also include anti-U1 RNP, associated syndromes. Anti-idiotypic antibod-

anti-Ro/SSA, anti-La/SSB, and anti-Sm. How- ies directed towards the binding site of on September 25, 2021 by guest. Protected copyright. ever, it is recognised that when autoimmune antibodies to viral tRNA might crossreact with myositis is associated with another CTD, the synthetases. Finally, a genetic background con- full criteria for the concomitant condition may ducive to the development of MSA is required. not be met.41 HLA associations have been established be- tween B8, DR3, DRw52, and tRNA synthetase tRNA synthetase syndrome syndrome.39 40 The tRNA synthetase syndrome is character- ised by myositis (PM or DM) (83–100%), Other myositis overlaps interstitial disease (50–80%), and Ray- SLE is associated with polymyositis in 4–16% naud’s phenomenon (60–93%).42 There is of cases and some reports have indicated that some doubt that this syndrome should be this overlap syndrome follows a benign course. regarded as a true overlap.43 The features of this However, a recent review of cases by Garton syndrome may be confused with those of SLE, and Isenberg48 has indicated that no significant scleroderma or RA44; however, tRNA syn- diVerence is seen between overlap and non- thetase syndrome is rarely observed as an over- overlap patients. Antibodies to Jo-1 were of low lap syndrome associated with other well prevalence in this group, although anti-56 kDa defined CTDs.45 46 tRNA synthetase syndrome nRNP antibodies were seen with high fre- is determined serologically by the presence of quency. Anti-56 kDa nRNP antibodies are antibodies against aminoacyl-tRNA syn- associated with myositis occurring together thetases, and diagnosis depends upon the with SLE; the presence of this antibody can detection of the appropriate autoantibody. The help to predict muscle involvement in this tRNA synthetases are a series of 20 cytoplas- group of patients.49 Other overlap associations mic enzymes that attach tRNA to its corre- with myositis have been reported but it is sponding amino acid during the assembly of doubtful that these reports reflect true overlap polypeptides. Antibodies against five diVerent syndromes rather than coincidence.

www.jclinpath.com Autoantibodies and overlap syndromes in autoimmune rheumatic disease 343

Table 3 Autoantibodies in scleroderma Scleroderma-polymyositis overlap

syndrome (PM-Scl) J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from Autoantibody Clinical importance The term “sclerodermatomyositis” was used Anti-DNA topoisomerase I (Scl-70) DiVuse scleroderma originally to define a group of patients with Anticentromere (proteins A, B C) Limited scleroderma features of both scleroderma and dermato- Anti-U3RNP (fibrillarin) DiVuse scleroderma—severe disease 52 Anti-7-2/8-2 RNP (Th/To) Limited scleroderma myositis. The main features of this syndrome Anti-RNA polymerases I, II, III DiVuse scleroderma, renal crisis are or myositis, , scleroderma- Anti-PM-Scl Scleroderma overlap like cutaneous changes, Raynaud’s phenom- Anti-U2 RNP Scleroderma overlap Anti-Ku Scleroderma overlap enon, and an association with specific auto- Antipyruvate dehydrogenase complex (M2) Limited scleroderma and scleroderma overlap antibodies: anti-PM-Scl, anti-Jo-1, anti-Ku, and anti-U2 RNP. The detection of specific auto- Scleroderma and overlap syndromes antibodies or autoantibody profiles assists accu- Scleroderma or systemic sclerosis is a general- rate diagnosis, allowing provision of optimal ised connective tissue disease involving sclerotic treatment. Patients with PM-Scl overlap re- changes in the skin and many other organ spond to small doses of corticosteroids and do systems. The condition encompasses a spec- not require intensive treatment; in contrast, trum of disorders that range from localised patients with scleroderma may not respond to forms such as morphoea with limited cutaneous such treatment.51 involvement (which may involve the internal Antibodies to PM-Scl prevail and are found 7853 organs after long periods) to diVuse cutaneous in 40–50% of patients. This antibody was disease invariably accompanied by early inter- first identified by double immunodiVusion and nal organ involvement.41 Patients with sclero- indirect immunofluorescence, where it pro- derma overlap are frequently referred to derma- duces a characteristic homogeneous nucleolar tologists owing to cutaneous symptoms and staining pattern together with weak staining of may be misdiagnosed as having scleroderma. the nucleoplasm. The PM-Scl antigen is Thus, to establish diagnosis and treatment for located at the site of ribosome assembly in the patients with scleroderma overlap syndromes, it granular component of the nucleolus, although is important that the associated clinical and its function is unknown; the PM-Scl complex 54 55 serological features are recognised. Clinically, may have a role in ribosomal maturation. It scleroderma has been described in association is a compound antigen consisting of up to 16 with SS, SLE, DM, Hashimoto’s thyroiditis, polypeptides with molecular weights ranging and primary biliary cirrhosis.41 50 Scleroderma from 20 to 110 kDa. Autoantibodies are directed predominantly against two molecules overlap syndromes are frequently referred to as 54 55 MCTDs; however, overlaps also occur in of 100 kDa (95%) and 75 kDa (50%). patients who do not have antibodies to nRNP However, although anti-PM-Scl antibodies are found in patients with PM-Scl overlap syn- and are associated with ANA of nucleolar stain- 7856 ing pattern; table 3 details the full range of drome in Europe and North America, they associated autoantibodies.51 are not found in Japanese patients. This is http://jcp.bmj.com/ probably the result of diVerences in genetic and environmental backgrounds.57 Anti-PM-Scl antibodies are strongly associated with HLA Raynaud’s phenomenon DR3; this HLA marker is found rarely in the Raynaud’s phenomenon is almost universal in Japanese population but is common in the nor- patients with scleroderma and is a feature of mal population of Europe and North America.

many other CTDs and overlap syndromes. Anti-Ku antibodies were described initially on September 25, 2021 by guest. Protected copyright. Table 4 shows the prevalence of Raynaud’s in Japanese patients with PM-Scl overlap, although they have since been described in phenomenon in these conditions. It is a disor- 58 der characterised by episodic, clearly demar- other patient groups. The target antigen is a cated, two or three phase colour change of the heterodimer of 70 kDa and 80 kDa proteins that acts as a regulatory subunit of DNA extremities (usually fingers), in response to dependent protein kinase, an enzyme that cold or emotion. Raynaud’s phenomenon catalyses the phosphorylation of nuclear pro- occurs in up to 5% of the normal population, teins. This enzyme complex is involved in tran- with over 90% of suVerers being female; the scription, DNA repair, antigen receptor V(D)J age of onset is usually below 25 years. Five per recombination, and apoptosis.57 These anti- cent of patients presenting with this condition bodies also present a nucleolar staining pattern eventually develop an autoimmune rheumatic by immunofluorescence. Other associated auto- disease. The best predictors of progression to antibodies include anti-DNA protein kinase CTD are an asymmetrical pattern of Ray- complex antibodies and anti-U2 RNP antibod- naud’s, positive ANA, and abnormal nail fold 41 ies, rare antibodies that usually coexist with capillary microscopy. anti-U1 RNP.8 The aetiology and pathogenic mechanisms Table 4 Prevalence of Raynaud’s phenomenon in of PM-Scl overlap remain unknown. A strong autoimmune rheumatic diseases (%) (adapted from association of anti-PM-Scl antibody with HLA Morrow et al)41 DR3 and an increased frequency of HLA DQA*0501 suggest a genetic influence.56 Scleroderma >95 Sjögren’s syndrome 20–50 However, these associations are not absolute Myositis 20–40 and it is probable that several independent fac- Systemic lupus erythematosus 20–30 tors can result in similar pathogenic mecha- Rheumatoid arthritis <5 nisms.59

www.jclinpath.com 344 Jury, D’Cruz, Morrow

Table 5 Autoimmune diseases associated with Sjögren’s Anti-Ro/SSA and anti-La/SSB antibodies are syndrome (adapted from Morrow et al)41 not specific for SS but are found commonly in J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from these patients (for a detailed review of these Rheumatoid arthritis 66 Scleroderma autoantibodies see Scofield and colleagues ). Systemic lupus erythematosus The Ro autoantigen is an RNP complex binding Polymyositis/dermatomyositis Primary biliary cirrhosis to the stem–loop structure of human cytoplas- Autoimmune hepatitis mic RNA (hYRNA). Antibodies directed Polyarteritis nodosa against two Ro polypeptides have been Diabetes Graves’ disease identified—52 kDa (Ro52) and 60 kDa (Ro60) Myasthenia gravis proteins—and there is evidence that the 52 kDa Coeliac disease protein binds to the 60 kDa protein, although this is not certain.67 The Ro 52 kDa protein has Other scleroderma overlap syndromes been isolated and characterised; it contains two An association between limited scleroderma zinc finger structures and a leucine zipper motif, (CREST syndrome), primary biliary cirrhosis, indicating a role in DNA and RNA binding and and SS has been reported.50 Patients are char- protein–protein interaction. La, which may bind acterised serologically by the presence of transiently to Ro, is an ATP dependent tran- autoantibodies directed against the pyruvate scription termination factor for RNA polymer- dehydrogenase complex (M2), centromere, ase III; it is a 48 kDa protein that binds polyU and Ro/SSA and/or La/SSB and genetically by stretches at the 3' end of unprocessed polymer- an association with HLA Cw6.60 ase III transcripts. Antibody responses to Ro, Scleroderma and RA may be seen as part of La, and other nuclear antigens, such as antipro- an overlap syndrome. Rheumatoid factor can teasomal antibodies, have been seen to vary between primary SS and SS overlap syn- be detected together with joint damage, but 68 69 whether this represents a distinct clinical entity dromes. Anti-Ro antibodies have been shown to fluctuate in parallel with disease activ- or part of the spectrum of scleroderma is 70 unknown.50 61 ity, although this is the exception rather than the rule in clinical practice. The relation between autoantibody profile and disease patho- SS overlap genesis is not clear; the penetration of these SS is a chronic inflammatory autoimmune dis- autoantibodies into living cells is powerful order aVecting the lachrymal and salivary evidence that antibodies to Ro and La are impli- glands and other exocrine organs throughout cated pathogenically in disease. For example, in the body. Dryness of eyes and mouth, chronic complete neonatal block,71 a possible fatigue, and are the most common mechanism of tissue damage includes the features of this disease, which may appear penetration of these autoantibodies into living solely, as primary SS, or together with other cells.69 In SS, autoantibodies are produced in

rheumatic disease as secondary SS. Sicca inflammatory lesions of the salivary glands and http://jcp.bmj.com/ syndrome describes a limited form of SS the immune response appears to be antigen restricted to dryness of eyes and mouth, and driven. Infection, resulting in mechanisms such this syndrome is frequently associated with as , reaction to “altered self”, other autoimmune disease but mainly SLE and intermolecular and intramolecular spreading RA (table 5). Moutsopoulos and Manous- might result in antibodies to multiple, linked, sakis62 have suggested additional classification self components frequently seen in SS.72 Viral

groups for patients with SS based on serologi- infection is strongly implicated as an aetiological on September 25, 2021 by guest. Protected copyright. cal and immunogenetic profiles. They have factor—for example, Epstein-Barr virus,73 retro- suggested that they can be divided into three viruses such as human leukaemia virus major subsets: those with anti-Ro/SSA and type 1,74 75 hepatitis C,76 and also organisms such La/SSB antibodies; those with no specific as Helicobacter pylori.77 HLA associations with autoantibody response; and those with anti- secondary SS have been identified; patients with bodies against autoantigens such as centro- anti-Ro52 and anti-Ro60 have an increased fre- mere, mitochondria, and thyroperoxidase. An quency of DR52, in contrast to associations extensive array of both organ and non-organ observed in patients with primary SS.78 Several specific autoantibodies may be detected in studies have reported an association between patients with SS (table 6), the most prominent HLA DQ and anti-Ro and anti-La antibodies.79 serological feature being the production of rheumatoid factor and ANA associated with Table 6 Organ specific and organ non-specific antibodies to Ro/SSA and La/SSB.63 However, autoantibodies associated with Sjögren’s syndrome overlap important diVerences in serological profiles syndromes (%) (adapted from Morrow et al)41 may be found, depending on the nature of the Antinuclear antibodies >1/80 (titre) 60–100 overlap syndrome partner. Anti-Ro/SSA 5–62 Rheumatoid factor is the most common Anti-La/SSB 5–42 Anti-Sm <5 autoantibody found in patients with SS. Anti-RNP 5–23 DiVerences between SS rheumatoid factor and Anti-dsDNA 5–56 that produced in patients with RA have been Rheumatoid factor 60–100 Antismooth muscle 30 seen, where rheumatoid factor variable genes Antisalivary duct 50–60 64 65 are close to germ line configurations in SS. Antithyroperoxidase 10–20 Occasionally, very high titres of rheumatoid Antigastric parietal cell 25–30 factor may produce a cryoglobulinaemic vascu- ds, double stranded; RNP, ribonuclear protein; Sm, Smith litis in these patients. antigen

www.jclinpath.com Autoantibodies and overlap syndromes in autoimmune rheumatic disease 345

Alternatively, HLA associations in secondary SS autoimmune target in PBC. PDC-E2 epitopes might reflect that of the associated disease. recognised by T cells have been isolated and J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from Splitting patients into subgroups based upon aberrant expression of MHC class II molecules, HLA associations, autoantibody profiles, and intercellular adhesion molecules, and lym- clinical features might provide insight into the phocyte costimulatory molecules has been pathogenesis of SS overlaps.62 described.84 Antibodies to nuclear antigens are found in up to 50% of patients with PBC and are RA/SLE overlap useful tools that aid the diagnosis of AMA nega- The existence of patients with a combination of tive PBC (autoimmune cholangitis). These features representing both RA and SLE is antibodies can also act as markers of overlaps acknowledged and termed ‘Rhupus’.41 80 Ar- with autoimmune rheumatic diseases and au- thropathy is associated with 5% of patients toimmune hepatitis (AIH).685 PBC specific with lupus and is typically non-erosive. Occa- ANA include antibodies to gp210, nucleoporin sionally, patients with SLE develop erosive p62, and Sp10082 and are detected by immun- arthritis with nodules suggesting an overlap ofluorescence. with RA. This is seen more frequently as patients become menopausal, when long PBC/CREST overlap standing SLE evolves into a more rheumatoid The association of PBC with limited cutaneous picture, with erosive joint disease. These systemic sclerosis or CREST syndrome is well conditions are rare and there is a paucity of documented.86 Patients present with concomi- information regarding autoantibody profiles tant features of both conditions, although it has and immunogenetics. It is debatable whether been suggested that patients with overlap these conditions represent true overlap, rather exhibit a mild form of PBC. Anticentromere than chance occurrence of both conditions in antibody (ACA) is seen in 50–80% of patients the same patient, or evolution of one disease with CREST syndrome, but is reported in into another.43 patients with other autoimmune disease, in- cluding up to 29% of patients with PBC.87 PBC overlap syndromes There are three major antigenic polypeptides Primary biliary cirrhosis (PBC), a chronic, recognised by ACA termed CENP-A, inflammatory, cholestatic disease of the liver CENP-B, and CENP-C, with an increased results in progressive and irreversible destruc- reactivity to CENP-C reported in PBC/ tion of small interlobular and septal bile ducts CREST overlap syndrome. However, a role for 81 these antibodies in disease pathogenesis has and liver failure. The principal treatment for 88 patients with PBC is ursodeoxycholic acid, not been ascertained. AMA are also seen in although advanced disease is treated by liver this overlap population; recently, antibodies to transplantation. The disease has an auto- the autoantigens E1â and E1á subunit of PDC immune aetiology and is frequently associated have been suggested as possible serological 82 83 indicators for the development of PBC in http://jcp.bmj.com/ with other autoimmune conditions (table 7). 83 89 The presence of antimitochondrial antibody patients with CREST syndrome. Weak (AMA) is a hallmark of PBC and these antibod- reactivity of almost all sera from patients and ies are found in 95% of patients. AMA are controls to the E3 subunit of PDC may shed present very early in the natural history of the some light on the mechanism by which disease, suggesting an intimate role in its patho- immunological tolerance is broken in PBC. An E3 component exists on the membrane of genesis. The major antigen termed M2 is 83 located on the inner mitochondrial membrane archaebacteria and eubacteria. Unidentified on September 25, 2021 by guest. Protected copyright. and is composed of at least five determinants. viral infection has been implicated in the The major autoantigen is the E2-subunit of the production of clonally expanded CD8 positive pyruvate dehydrogenase complex (PDC-E2). T cells in these patients; the function of these T However, patients with PBC frequently display cell clones remains unknown but a role in pathogenesis is possible.90 antibodies to other mitochondrial antigens, such 88 as the E1á and E1â subunits of PDC, protein X Akimoto and colleagues showed that 75– of the PDC, and the E2 component of the 91% of patients with PBC/CREST overlap syndrome display features of SS. An overlap of branched chain á-ketoacid dehydrogenase and 91 á-ketogluterate dehydrogenase complex. The SS with PBC alone has also been reported. detection of AMA is the most important diagnostic test for PBC, although these antibod- Autoimmune hepatitis (AIH) overlap ies do not appear to contribute directly to the syndromes pathogenesis of the disease.82 Recent reports AIH is associated with a range of autoantibodies have suggested that biliary epithelial cells are the including anti-SMA, ANA, anti-LKM, and anti-asialoglycoprotein receptor. Variant forms Table 7 Autoimmune syndromes associated with primary of AIH are common, including combinations of biliary cirrhosis AIH, PBC, primary sclerosing cholangitis,92 or chronic viral hepatitis.93 A detailed discussion of Myasthenia gravis Autoimmune thyroid disease autoimmune hepatitis is beyond the scope of this Mixed connective tissue disease review (for an overview of associated autoanti- Systemic lupus erythematosus bodies see Manns,6 Czaja,94 Chazouilleres et al,95 Pernicious anaemia 96 Polymyositis and Lohse and colleagues ). Limited scleroderma (CREST) Liver disease is also recognised as an impor- Rheumatoid arthritis tant but very rare clinical problem in some Sjögren’s syndrome patients with SLE, so called lupoid hepatitis.

www.jclinpath.com 346 Jury, D’Cruz, Morrow

Satoh and colleagues97 reported a new auto- 9 Fine RM. Overlap syndromes. Int J Dermatol 1991;30:479– antibody to a tRNA related antigen that 80. 10 Isenberg DA, Black C. Naming names! Comment on the J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from appears to be specific for a clinical subset of article by Smolen and Steiner. Arthritis Rheum 1999;42: 191–6. patients who have a mild form of SLE 11 Brand CA, Rowley MJ, Tait BD, et al. Coexistent associated with AIH. rheumatoid arthritis and systemic lupus erythematosus: clinical, serological and phenotypic features. Ann Rheum Primary sclerosing cholangitis and AIH, Dis 1992;51:173–6. both chronic liver diseases with a probable 12 Horiki T, Moriuchi J, Takaya M, et al. The coexistence of systemic sclerosis and rheumatoid arthritis in five patients: autoimmune background, have been reported clinical and immunogenetic features suggest a distinct to appear as an overlap. Patients fill the entity. Arthritis Rheum 1996;39:152–6. 13 Nimelstein SH, Brody ST, McShane D, et al. Mixed diagnostic criteria for both conditions and connective tissue disease: a subsequent evaluation of the autoantibodies associated with both conditions original 25 patients. Medicine (Baltimore) 1980;59:239–48. 14 LeRoy EC, Maricq HR, Kalaleh MB. UndiVerentiated con- are detected; namely, ANA and anti-SMA nective tissue syndromes. Arthritis Rheum 1980;23:341–3. associated with AIH, together with perinuclear 15 Black C, Isenberg D. Mixed connective tissue disease— goodbye to all that. Br J Rheumatol 1993;32:348–9. antineutrophil cytoplasmic antibodies associ- 16 Williams HJ, Alarcon GS, Neuner R, et al. Early undiVeren- ated with primary sclerosing cholangitis.95 tiated connective tissue disease. V. An inception cohort 5 years later: disease remissions and changes in diagnoses in well established and undiVerentiated connective tissue dis- Conclusion eases. J Rheumatol 1998;25:261–8. In a clinical setting, autoantibodies are used to 17 Alarcon GS. Unclassified or undiVerentiated connective tis- sue disease. Baillieres Best Practice and Research in Clinical establish diagnosis, estimate prognosis, follow Rheumatology 2000;14:125–37. disease progression, and monitor treatment 18 Sharp GC, HoVman DO. Clinical, immunologic and immunogenetic evidence that mixed connective tissue dis- regimens. Some are specific disease markers ease is a distinct entity: comment on the article by Smolen and play a key role in recognised diagnostic and Steiner. Arthritis Rheum 1999;42:190–1. 19 Van den Hoogen FHJ, van de Putte LBA. Is mixed connec- criteria. In some cases, it is the profile of tive tissue disease a myth? In: Isenberg DA, Tucker LB, eds. autoantibodies, together with other clinical Controversies in rheumatology. London: Dunitz, 1997:87–95. 20 Maddison PJ. Mixed connective tissue disease: overlap syn- features, that aids diagnosis. dromes. Baillieres Best Practice and Research in Clinical The identification of overlapping clinical Rheumatology 2000;14:111–24. 21 Smolen JS, Steiner G. Mixed connective tissue disease: to be features in a given patient is important because or not to be? Athritis Rheum 1998;41:768–77. treatment might need to be directed specifi- 22 Sharp GC, Irvin WS, Tan EM, et al. Mixed connective tissue disease: an apparently distinct rheumatic disease syndrome cally at some of these features. The develop- associated with a specific antibody to an extractable nuclear ment of myositis in a patient with apparent antigen. Am J Med 1972;52:148–59. 23 Tan EM. Antinuclear antibodies: diagnostic markers for scleroderma is a useful example where the autoimmune diseases and probes for cell biology. Adv myositis may respond to active treatment. It is Immunol 1989;52:148–51. 24 Alarcon-Segovia D, Villarreal M. Classification and diag- in overlap patients that autoantibody profiles nostic criteria for mixed connective tissue diseases. In: and possibly genetic associations might be Kasukawa R, Sharp GC, eds. Mixed connective tissue diseases and anti-nuclear antibodies. Amsterdam: Elsevier, 1987:33– most useful in predicting response to treatment 40. and long term prognosis. 25 Kasukawa R, Tojo T, Miyawaki S, et al. Preliminary diagnostic criteria for classification of mixed connective tis- Approximately 50 predominantly nuclear sue disease. In: Kasukawa R, Sharp GC, eds. Mixed connec-

proteins act as autoantigens in autoimmune tive tissue diseases and anti-nuclear antibodies. Amsterdam: http://jcp.bmj.com/ Elsevier, 1987:41–7. rheumatic diseases; most autoantigens are 26 Steiner G, Skriner K, Smolen JS. Autoantibodies to the A/B multiprotein or nucleoprotein complexes with proteins of the heterogeneous nuclear ribonucleoprotein complex: novel tools for the diagnosis of rheumatic important functions in cell regulation, such as diseases. Int Arch Immunol 1996;111:314–19. RNA splicing. The basis of pathogenesis is 27 Hoet RM, Koornneef DJ, De Rooij LB, et al. Changes in anti-U1 RNA antibody levels correlate with disease activity important to determine because it becomes the in patients with systemic lupus erythematosus overlap syn- focus for possible therapeutic intervention. drome. Arthritis Rheum 1992;35:1202–9. Characterisation of antigens recognised by 28 Komatireddy GR, Wang GS, Sharp GC, et al. Antiphos- pholipid antibodies among anti-U1–70kDa autoantibody on September 25, 2021 by guest. Protected copyright. autoantibodies associated with non-organ spe- positive patients with mixed connective tissue disease. J cific autoimmune disease, together with the use Rheumatol 1997;24:319–22. 29 Hassfeld W, Steiner G, Studnicka A, et al. Autoimmune of molecular biology technology, will help to response to the spliceosome: an immunologic link between clarify similarities and di erences between rheumatoid arthritis, mixed connective tissue disease and V systemic lupus erythematosus. Arthritis Rheum 1995;38: various CTDs and overlap syndromes. This 777–85. 30 Klein Gunnewiek JMT, Van De Putte LBA, Van Venrooij might lead to the development of new diagnos- WJ. The U1 snRNP complex: an autoantigen in connective 98 99 tic and therapeutic strategies. tissue diseases. An update. Clin Exp Rheumatol 1997;15: 549–60. 31 Margaux J, Hayem G, Palazzo E, et al. Clinical usefulness of The authors thank Professor D Isenberg for his advice and autoantibodies to U1snRNP proteins in mixed connective support. tissue disease and systemic lupus erythematosus. Rev Rhum Engl Ed 1998;65:378–86. 1 Tan EM, Cohen AS, Fries JF, et al. The 1982 revised crite- 32 Hoet RMA, PieVers M, Stassen MHW, et al. The ria for the classification of systemic lupus erythematosus. importance of the light chain for the epitope specificity of Arthritis Rheum 1982;25:1271–5. human anti-U1 small nuclear RNA autoantibodies present 2 Arnett FC, Edworthy SM, Bloch DA, et al. The American in systemic lupus erythematosus patients. J Immunol 1999; Rheumatism Association 1987 revised criteria for the 163:3304–12. classification of rheumatoid arthritis. Arthritis Rheum 1987; 33 Burdt MA, HoVman RW, Deutscher SL, et al. Long term 31:315–24. outcome in mixed connective tissue disease: longitudinal 3 Johnson PJ, McFarlane IG. Meeting report of the inter- clinical and serologic findings. Arthritis Rheum 1999;42: national autoimmune hepatitis group. 1993;18: 899–909. 998–1005. 34 Steiner G, Skriner K, Hassfeld W, et al. Clinical and 4 Cervera R, Khamashta MA, Hughes GVR. Overlap immunological aspects of autoantibodies to RA33/hnRNP- syndromes. Ann Rheum Dis 1990;49:947–8. A/B proteins—a link between RA, SLE and MCTD. Mol 5 Maddison PJ. Mixed connective tissue disease, overlap syn- Biol Rep 1996;23:167–71. dromes and . Ann Rheum Dis 1991;50: 35 Barakat S, Briand JP, Abuaf N, et al. Mapping of epitopes on 887–93. U1 sn RNP polypeptide A with synthetic peptides and 6 Manns MP. Recent developments in autoimmune liver dis- autoimmune sera. Clin Exp Immunol 1991;86:71–8. eases. J Gastroenterol Hepatol 1997;12(suppl):S256–71. 36 Hassan AB, Ronnelid J, Gunnarsson I, et al. Increased 7 Mimori T. Structures targeted by the immune system in serum levels of immunoglobulins, C-reactive protein, type myositis. Curr Opin Rheumatol 1996;8:521–7. 1 and type 2 cytokines in patients with mixed connective 8 Mimori T. Autoantibodies in connective tissue diseases: tissue disease. J Autoimmun 1998;11:503–8. clinical significance and analysis of target autoantigens. 37 HoVman RW,Sharp GC, Deutscher SL. Analysis of anti-U1 Intern Med 1999;38:523–32. RNA antibodies in patients with connective tissue disease:

www.jclinpath.com Autoantibodies and overlap syndromes in autoimmune rheumatic disease 347

association with HLA and clinical manifestations of lupus erythematosus and Sjogren’s syndrome: a prospec- disease. Arthritis Rheum 1995;38:1837–44. tive study. Clin Exp Rheumatol 1999;17:63–8.

38 Mimori T. Structures targeted by the immune system in 71 Buyon JP. Neonatal lupus syndromes. Curr Opin Rheumatol J Clin Pathol: first published as 10.1136/jcp.54.5.340 on 1 May 2001. Downloaded from myositis. Curr Opin Rheumatol 1996;8:521–27. 1994;6:523–9. 39 TargoV IN. Immune manifestations of inflammatory muscle 72 Topfer F, Gordon T, McCluskey J. Intra- and intermolecu- disease. Rheum Dis Clin North Am 1994;20:857–80. lar spreading of autoimmunity involving the nuclear self 40 TargoV IN. Autoantibodies in polymyositis. Rheum Dis Clin antigens La (SS-B) and Ro (SS-A). Proc Natl Acad Sci USA North Am 1992;18:455–82. 1995;92:875–79. 41 Morrow WJW, Nelson JL, Watts RA, et al. Sjögren’s 73 Venables PJW. Polymyositis-associated overlap syndromes. syndrome. In: Autoimmune rheumatic disease, 2nd ed. Br J Rheumatol 1996;35:305–6. Oxford: Oxford University Press, 1999:147–241. 74 Tominaga M, Mitita K, Nakamura H, et al. Expression of 42 Marguerie C, Bunn CC, Copier J, et al. Polymyositis, metalloproteinase-2 (gelatinase A) in labial salivary glands pulmonary fibrosis and autoantibodies to aminoacyl-tRNA of patients with Sjogren’s syndrome with HTLV-1 synthetase. 1990; :1019–38. QJMed 77 infection. Clin Exp Rheumatol 1999;17:463–6. 43 Alarcon-Segovia D. Mixed connective tissue disease and 75 Hida A, Kawabe Y, Kawakami A, . HTLV-1 associated overlap syndromes. 1994; :309–16. et al Clin Dermatol 12 Sjogren’s syndrome is aetiologically distinct from anti- 44 Miller FW. Myositis specific antibodies: touchstones for understanding the inflammatory . 1993; centromere antibodies positive Sjogren’s syndrome. Ann JAMA 1999; :320–2. 270:1846–9. Rheum Dis 58 45 O’Neill TW, Maddison PJ. Rheumatoid arthritis associated 76 Ramos-Casals M, Garcia-Carrasco M, Cervera R, et al. Sjogren’s syndrome and hepatitis C virus. with myositis and anti-Jo-1 antibody. J Rheumatol 1993;20: Clin Rheumatol 141–3. 1999;18:93–100. 46 Love LA, LeV RL, Fraser DD, et al. A new approach to the 77 Aragona P, Magazzu G, Macchia G, et al. Presence of anti- classification of idiopathic inflammatory : bodies against Helicobacter pylori and its heat-shock myositis-specific autoantibodies define useful homogene- protein 60 in the serum of patients with Sjogren’s ous patient groups. Medicine (Baltimore) 1991;70:360–74. syndrome. J Rheumatol 1999;26:1306–11. 47 Plotz PH, Autoantibodies are anti-idiotype antibodies to 78 Wilson RW, Provost TT, Bias WB, et al. Sjogren’s syndrome. antiviral antibodies. Lancet 1983;ii:824–6. Influence of multiple HLA-D region alloantigens on clini- 48 Garton MJ, Isenberg DA. Clinical features of lupus myositis cal and serologic expression. Arthritis Rheum 1984;27: versus idiopathic myositis: a review of 30 cases. Br J Rheu- 1245–53. matol 1997;36:1067–74. 79 Harley JB, Reichlin M, Arnett FC, et al. Gene interaction at 49 Ioannou Y, Sultan S, Isenberg DA. Myositis overlap HLA-DQ enhances autoantibody production in primary syndromes. Curr Opin Rheumatol 1999;11:468–74. Sjogren’s syndrome. Science 1986;232:1145–7. 50 Bennett RM. Scleroderma overlap syndromes. Rheumatic 80 Hamasaki K, Mimura T, Kanda H, et al. Development of Dis Clin North Am 1990;16:185–98. systemic lupus erythematosus in a rheumatoid arthritis 51 Jablonska S, Chorzelski TP, Blaszczyk M, et al. Scleroderma/ patient with anti-ribosomal P protein antibody. Lupus polymyositis overlap syndromes and their immunologic 1997;6:734–6. markers. Clin Dermatol 1993;10:443–56. 81 Kaplan MM. Primary biliary cirrhosis. N Engl J Med 1987; 52 TuVanelli DL, Winkelmann RK. . A 316:521–8. clinical study of 727 cases. Arch Dermatol 1961;84:359–71. 82 Strassburg CP, Jaeckel E, Manns MP. Anti-mitochondrial 53 Reichlin M, Maddison P, TargoV I, et al. Antibodies to a antibodies and other immunological tests in primary biliary nuclear/nucleolar antigen in patients with polymyositis cirrhosis. Eur J Gastroenterol Hepatol 1999;11:595–601. overlap syndromes. J Clin Immunol 1984;4:40–4. 83 Akimoto S, Ishikawa O, Takagi H, et al. Immunological fea- 54 Bluthner M, Bautz EKF, Bautz FA. Mapping of epitopes rec- tures of patients with primary biliary cirrhosis (PBC) over- ognised by PM/Scl autoantibodies with gene-fragment phage lapping systemic sclerosis: a comparison with patients with display libraries. J Immunol Methods 1996;198:187–98. PBC alone. J Gastroenterol Hepatol 1998;13:897–901. 55 Alderuccio F, Chan EKL, Tan EM. Molecular characteriza- 84 Joplin RE, Neuberger JM. Immunopathology of primary tion of an autoantigen of PM/Scl in the polymyositis/ biliary cirrhosis. Eur J Gastroenterol Hepatol 1999;11:587– scleroderma overlap syndrome. J Exp Med 1991;173:941– 93. 52. 85 Czaja AJ. Frequency and nature of the variant syndromes of 56 Hausmanowa-Petrusewicz I, Kowalska-Oledzka E, Miller autoimmune liver disease. Hepatology 1998;28:360–5. FW, et al. Clinical, serologic and immunogenetic features in 86 Powell FC, Schroeter AL, Dickson ER. Primary biliary cir- Polish patients with idiopathic inflammatory myopathies. rhosis and the CREST syndrome: a report of 22 cases. QJ Arthritis Rheum 1997;40:1257–66. Med 1987;62:75–82. 57 Yamanishi Y, Maeda H, Katayama S, . Scleroderma- et al 87 Tubach F, Hatem G, Elias A, et al. Anticentromere antibod- http://jcp.bmj.com/ polymyositis overlap syndrome associated with anti-Ku ies in rheumatologic practice are not consistently associ- antibody and rimmed vacuole formation. J Rheumatol ated with scleroderma. Rev Rhum Engl Ed 1997;64:362–7. 1996;23:1991–4. 88 Akimoto S, Ishikawa O, Muro Y et al. Clinical and immuno- 58 Mimori T, Akizuki M, Yamagata H, et al. Characterization logical characterisation of patients with systemic sclerosis of a high molecular weight acidic nuclear protein overlapping primary biliary cirrhosis: a comparison with recognized by autoantibodies in sera from patients with patients with systemic sclerosis alone. J Dermatol 1999;26: polymyositis-scleroderma overlap. J Clin Invest 1981;68: 18–22. 611–20. 89 Fujimoto M, Sato S, Ihn Hironobu, et al. Autoantibodies to 59 Bunn CC, Black CM. Systemic sclerosis: an autoantibody pyruvate dehydrogenase complex in patients with systemic mosaic. Clin Exp Immunol 1999;117:207–8. sclerosis: possible role of anti-E1á antibody as a serologic 60 Akiyama Y, Tanaka M, Takeishi M, . Clinical, serologi- et al indicator for development of primary biliary cirrhosis. on September 25, 2021 by guest. Protected copyright. cal and genetic study in patients with CREST syndrome. Arthritis Rheum 1995;38:985–9. 2000; :451–6. Intern Med 39 90 Mayo MJ, Jenkins RN, Combes B, et al. Association of clon- 61 Horiki T, Moriuchi J, Takaya M, et al. The coexistence of ally expanded T cells with the syndrome of primary biliary systemic sclerosis and rheumatoid arthritis in five patients: cirrhosis and limited scleroderma. Hepatology 1999;29: clinical and immunogenic features suggest a distinct entity. 1635–42. Arthritis Rheum 1996;39:152–6. 91 Fujikura S, Davis P, Prindiville T, et al. Sjogren’s syndrome 62 Moutsopoulos HM, Manoussakis MN. Lumping or split- and primary biliary cirrhosis: presence of autoantibodies to ting autoimmune rheumatic disorders? Lessons from purified mitochondrial 2-OXO acid dehydrogenates. Sjogren’s syndrome. 1998; :1263–4. J Br J Rheumatol 37 1990;17:1453–7. 63 Von Muhlen CA, Tan EM. Autoantibodies in the diagnosis Rheumatol of systemic rheumatic diseases. 92 Gohlke F, Lohse AW, Dienes HP, et al. Evidence for an over- Semin Arthritis Rheum lap syndrome of autoimmune hepatitis and primary sclero- 1995;24:323–58. sing cholangitis. J Hepatol 1996;24:699–705. 64 Elagib KE, Borretzen M, Jonsson R, et al. Rheumatoid factors in primary Sjogren’s syndrome (pSS) use diverse VH region 93 Klein R, Zanger UM, Berg T, et al. Overlapping but distinct genes, the majority of which show no evidence of somatic specificities of anti-liver-kidney microsome antibodies in hypermutation. Clin Exp Immunol 1999;117:388–94. autoimmune hepatitis type II and hepatitis C revealed by recombinant native CYP2D6 and novel peptide epitopes. 65 Elagib KEE, Tengner PIA, Levi M, et al. Immunoglobulin variable genes and epitope recognition of human mono- Clin Exp Immunol 1999;118:290–7. clonal anti-Ro 52-kd in primary Sjogrens syndrome. 94 Czaja AJ. The variant forms of autoimmune hepatitis. Ann Arthritis Rheum 1999;42:2471–81. Intern Med 1996;125:588–98. 66 Scofield RH, Farris AD, Horsfall AC, et al. Fine specificity 95 Chazouilleres O, Wendum D, Serfaty L, et al.Primary of the autoimmune response to the Ro/SSA and La/SSB biliary cirrhosis-autoimmune hepatitis overlap syndrome: ribonucleoproteins. Arthritis Rheum 1999;42:199–209. clinical features and response to . Hepatology 1998; 67 Kelekar A, Saitta MR, Keene JD. Molecular composition of 28:296–301. Ro small ribonucleoprotein complexes in human cells: 96 Lohse AW, Meyer zum Buschenfelde KH, Franz B, et al. intracellular localisation of the 60- and 52kD proteins. J Characterization of the overlap syndrome of primary Clin Invest 1994;93:1637–44. biliary cirrhosis (PBC) and autoimmune hepatitis: evi- 68 Ricchiuti V, Briand JP, Meyer O, et al. Epitope mapping with dence for it being a hepatic form of PBC in genetically sus- synthetic peptides of 52 kDa SSA/Ro protein reveals ceptible individuals. Hepatology 1999;29:1078–84. heterogeneous antibody profiles in human autoimmune 97 Satoh T, Hirakata M, Yoshida T, et al. Systemic lupus sera. Clin Exp Immunol 1994;95:397–407. erthyematosus associated with autoimmune hepatitis: two 69 Feist E, Kuckelkorn U, Dorner T, et al. Autoantibodies in cases with novel autoantibodies to transfer RNA-related primary Sjogren’s syndrome are directed against proteaso- antigens. Clin Rheumatol 1997;16:305–9. mal subunits of the á and â type. Arthritis Rheum 1999;42: 98 Reeves WH, Satoh M. Features of autoantigens. Mol Biol 697–702. Rep 1996;23:217–26. 70 Praprotnik S, Bozic B, Kveder T, et al. Fluctuation of 99 Fritzler MJ. Autoantibodies: diagnostic fingerprints and anti-Ro/SS-A antibody levels in patients with systemic etiologic perplexities. Clin Invest Med 1997;20:50–6.

www.jclinpath.com