If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The progressively growing knowledge of the pathophysiology of a number of immune-mediated gastrointestinal and liver disorders, including autoimmune atrophic gastritis, coeliac disease, autoimmune enteropathy, inflammatory bowel disease, autoimmune hepatitis, primary sclerosing cholangitis, primary biliary cholangitis and autoimmune pancreatitis, together with the improvement of their detection methods have increased the diagnostic power of serum antibodies. In some cases – coeliac disease and autoimmune atrophic gastritis – they have radically changed gastroenterologists’ diagnostic ability, while in others – autoimmune hepatitis, inflammatory bowel disease and autoimmune pancreatitis – their diagnostic performance is still inadequate. Of note, serum antibody misuse in clinical practice has raised a number of controversies, which may generate confusion in the diagnostic management of the aforementioned disorders. In this review, we critically re-evaluate the usefulness of serum antibodies as biomarkers of immune-mediated gastrointestinal and liver disorders, and discuss their pitfalls and merits.
The expansion of the pathogenic knowledge of immune-mediated gastrointestinal and liver diseases, including autoimmune atrophic gastritis (AAG), coeliac disease (CoeD), inflammatory bowel disease (IBD), autoimmune hepatitis (AIH), primary sclerosing cholangitis (PSC), primary biliary cholangitis (PBC) and autoimmune pancreatitis (AIP), has led to a deeper awareness of the clinical management of these conditions [
]. In particular, serum antibodies have increased our diagnostic power in most of the aforementioned disorders (Table 1, Table 2), although their misuse has often led to diagnostic mistakes thus generating a number of controversies regarding their accuracy and appropriate use in real life. Fig. 1 shows the accuracy of serum antibodies according to the figures commonly reported in the literature.
Table 1Serum antibody biomarkers of immune-mediated gastrointestinal disorders.
Fig. 1Schematic representation of the accuracy of the most commonly used serum antibodies in the diagnosis of immune-mediated gastrointestinal and liver diseases. Accuracy is calculated as the average of sensitivity and specificity by using the best available tests in the literature. AEA, anti-enterocyte antibody; AGA, anti-gliadin antibody; AMA, anti-mitochondrial antibody; ANA, anti-nuclear antibody; ASCA, anti-Saccharomyces cerevisiae antibody; DGP, anti-deamidated gliadin peptide antibody; EMA, anti-endomysial antibody; GCA, anti-goblet cell antibody; IFA, anti-intrinsic factor antibody; Ig, immunoglobulin; LC1, anti-liver cytosol 1 antibody; LKM1, anti-liver kidney microsomal antibody 1; pANCA, perinuclear anti-neutrophil cytoplasmic antibody; PCA, anti-gastric parietal cell antibody; SLA/LP, anti-soluble liver antigen/liver-pancreas antibody; TTA, anti-tissue transglutaminase antibody.
On this basis, we aimed to critically re-evaluate the usefulness of serum antibodies as biomarkers in immune-mediated gastrointestinal and liver disorders, and to discuss both their pitfalls and merits. For clarity, we decided to divide the review and discussion by diseases instead of by biomarkers, even if serum antibodies are the core of the review.
2. Autoimmune atrophic gastritis
AAG is an organ-specific disease which affects the corpus-fundus mucosa of the stomach, causing hypo-achlorhydria, deficiency of vitamin B12 and iron [
], and anti-parietal cell antibodies (PCA) and anti-intrinsic factor antibodies (IFA) could be potentially helpful in the diagnosis of this condition (Table 1) [
], and they can be identified by either immunofluorescence or enzyme-linked immunosorbent assay (ELISA), the latter being more accurate than the former [
]. We recently followed-up 58 patients with serum PCA positivity and normal gastric mucosa, and we observed that thirteen of them subsequently developed atrophy over a median 30-month follow-up period [
]. Of course, further studies are needed to verify whether the isolated PCA positivity could be considered a hallmark of “potential AAG”. A few considerations regarding the value of PCA in clinical practice should be made. Firstly, there is evidence from a single study [
] that serum PCA titres may fluctuate along the natural course of AAG, becoming negative in patients with longstanding disease, thus affecting PCA accuracy in late AAG. In these cases, if the clinical suspicion is high, other laboratory parameters and histology should be taken into account before ruling out AAG diagnosis. In particular, in a previous study we designed and validated a laboratory score model that evaluates serum basal 17-gastrin, haemoglobin and mean cell volume, able to detect AAG with a high accuracy [
Gastric parietal cell antibodies, Helicobacter pylori infection, and chronic atrophic gastritis: evidence from a large population-based study in Germany.
Cancer Epidemiol Biomarkers Prev.2013; 22: 821-826
]. Although there are no studies assessing the performance of PCA in a case-finding strategy, the relevant association of AAG with other autoimmune disorders [
] might make it worth using serology to screen patients with autoimmune thyroiditis, type I diabetes, Addison’s disease and vitiligo. In addition, the value of PCA in patients with megaloblastic anaemia has been widely investigated and confirmed [
]. IFA, even detected by ELISA, due to their low sensitivity (<30%) have limited clinical usefulness. However, in patients presenting with pernicious anaemia, a late finding in AAG [
The discovery of serum antibodies specific for CoeD is certainly the key factor that revolutionized our knowledge of this condition (Table 1). Formerly, CoeD was considered to be a rare disease, affecting almost exclusively children, and to be taken into account only in the case of severe malabsorption symptoms. Thanks to “coeliac serology” we know that CoeD is very frequent, it affects adults as much as children, and its clinical presentation is extremely variable [
]. They were detected by standard indirect immunofluorescence on cryostat section of rodent tissues, stomach, liver and kidney. Although some authors reported remarkably good levels of sensitivity and specificity [
], in everyday clinical practice the results were more disappointing and they never became a test used world-wide. The era of coeliac serology kicked off in the early 1980s with the discovery of ELISA anti-gliadin antibodies (AGA) and immunofluorescent anti-endomysial antibodies (EMA) [
]. In 1997, the discovery of the role of tissue transglutaminase 2 (TG2) in the pathogenesis of CoeD made it possible not only to detect EMA by means of an ELISA technique, i.e. anti-tissue transglutaminase antibodies (TTA), but also to detect deamidated gliadin peptide antibodies (DGP) [
]. A total of seven different types of coeliac antibodies have been investigated so far. However, it must be stressed that nowadays only EMA, TTA and DGP are the antibodies that need to be taken into account. Anti-reticulin and AGA are now obsolete, have exhausted their role and should be abandoned. Anti-actin and anti-glutenin antibodies never managed to get beyond the experimental level and enter into clinical practice, although IgA anti-actin antibodies were shown to correlate with histological damage.
EMA are directed against TG2 in soft connective tissues surrounding smooth muscle fibres, the so-called endomysium. They are detected with traditional immunofluorescence on cryostat sections of monkey oesophagus, monkey jejunum or human umbilical cord. When detected on monkey jejunum they are defined jejunum antibodies. Although the very first studies clearly showed that EMA of IgA class are the relevant ones and IgG EMA should be taken into account only in patients with IgA deficiency [
], it is nowadays quite common to have patients tested for both IgA and IgG EMA “in case there is an unknown IgA deficiency”. The problem with this strategy is that the sensitivity and specificity of IgG EMA in subjects with normal IgA levels are rather low. So, immunocompetent subjects with false positive IgG EMA are more common than patients with unknown IgA deficiency and true positive IgG EMA. Therefore, testing for total IgA should be part of the serological search for CoeD, and only when total IgA deficiency is detected should IgG EMA be performed [
]. The sensitivity of EMA has been reported to be around 95%, and since EMA specificity is close to 100%, a positive patient with a normal duodenal biopsy implies a diagnosis of potential CoeD [
]. Therefore, although they are defined with two different names, EMA and TTA are exactly the same antibodies. Again, only TTA of IgA class have a diagnostic role and TTA of IgG class should be considered only in the case of IgA deficiency. An ELISA technique is obviously cheaper to perform and does not require specific training for the operator. However, we do not agree with those authors who suggest that EMA detection is subjective and operator-dependent, and that TTA are more reliable because they provide a numerical result [
]. An EMA positive pattern is not just “stronger” than a negative one but it is different. We could say that to distinguish positive EMA from negative ones it is like distinguishing the Eiffel tower from the leaning tower of Pisa: a glance is enough and we do not need any measurement, any number to distinguish between them!
The sensitivity of TTA is maybe even higher than that of EMA but the specificity is not 100%. The main problem is represented by “low positive” TTA that are often unrelated with CoeD. We showed a few years ago that there is a very good relationship between TTA optical density and EMA titres, though this relationship tends to be weaker at low EMA titres and at low TTA optical density [
], we strongly recommend that weak TTA should always be confirmed by EMA. In children, even strong TTA need to be confirmed by EMA according to the latest ESPGHAN criteria, which allow the diagnosis of CoeD without duodenal biopsy. This can be done providing that TTA are higher than 10 times the normal value, EMA are found positive in a second serum sample, HLA-DQ2 and/or DQ8 are positive, clear-cut symptoms consistent with CD are present, and there is a response to a gluten-free diet [
]. Finally, the observation that during an infectious disease TTA can be produced temporarily and independently of gluten further supports the importance of confirming TTA-positivity with EMA [
As regards DGP, similarly to old AGA, both DGP of IgG and IgA class should be taken into account. Since the sensitivity of EMA/TTA tends to be lower in children below 3 years of age, DGP have been proposed as the serological tool of choice in this age group [
The proper use of coeliac antibodies in adulthood should be based on the pre-test probability of CoeD. Screening of the general population, i.e. mass screening, should be performed with TTA first, followed by EMA on TTA-positive samples and then duodenal biopsy. In at risk groups with an expected prevalence of CoeD ranging from 5 to 10% (case-finding strategy), patients should undergo either EMA or TTA determination followed by duodenal biopsy in those who test positive. Finally, patients with frank symptoms of malabsorption and unexplained malnutrition should undergo upper endoscopy with duodenal biopsies in spite of the EMA/TTA result [
Tons of papers have been written on the use of coeliac antibodies in the follow-up of CoeD and for checking gluten-free diet adherence. Although they are reported as a satisfactory tool in some papers [
]. Conversely, we showed that coeliac antibodies can persist despite strict dietary adherence and good histological response. More precisely, when histological response was used as the gold standard, the sensitivity and specificity of the serological response was 48% and 71%, respectively [
Common variable immunodeficiency is a condition that can present subtotal villous atrophy that does not always respond to a gluten-free diet. So, proving that these patients are also affected by CoeD can be a real challenge. We have recently shown that CoeD serology has no role in these patients and can even be misleading [
]. Therefore, the histological response to a gluten-free diet, together with the presence of DQ2 or DQ8 molecules, remains the only diagnostic criterion for CoeD in these patients.
4. Autoimmune enteropathy
Autoimmune enteropathy (AIE) is a chronic enteropathy first described in children [
], the diagnosis had to be based on subtotal villous atrophy refractory to any dietary exclusion, anti-enterocyte antibodies (EA) and/or associated autoimmune conditions, and absence of significant immunodeficiency. EA are therefore the serological marker of this condition (Table 1). They are detected by indirect immunofluorescence on cryostat sections of monkey or human jejunum where they bind to the cytoplasm and, to a lesser extent, the brush border of villi and crypt enterocytes (Fig. 2).
Fig. 2Strong positivity of anti-enterocyte antibodies (white arrows) detected by indirect immunofluorescence on cryostat sections of monkey jejunum (brush border and enterocyte cytoplasm). Original magnification: ×250.
]. Since then we have been actively looking for an AIE that represents an important differential diagnosis with CoeD and its complications. Although we are a referral centre for CoeD and we use monkey jejunum on a regular basis to look for EMA, only three more patients have been diagnosed in the last 20 years. The rarity of the condition is therefore the first obstacle to estimate the accuracy of EA for AIE, but it is not the only one. A second problem is linked to the diagnostic criteria themselves. Although those originally proposed by Unsworth and Walker-Smith [
] work perfectly well for children, we are not sure that this is also true for adults. In children, refractory CoeD simply does not exist, and so there is no problem of differential diagnosis with AIE. This is obviously not the case in adults where refractoriness and other complications of CoeD are a major problem. Since associated autoimmune conditions are very frequent in CoeD, EA are the only tool that can be used to discriminate between adult AIE and refractory CoeD. So, sensitivity of EA would inevitably be 100%. After having tested serum samples from more than 2200 patients undergoing duodenal biopsy, we think that the specificity of EA for AIE is virtually 100%. In children, in whom the diagnosis does not require positive EA, the highest sensitivity (91%) was obtained in 12 children affected by immune dysregulation, polyendocrinopathy, enteropathy, or X-linked disease, which is a syndromic form of AIE [
]. However, EA specificity in AIE is still under debate, EA being said to be positive in more than two-thirds of human immunodeficiency virus-infected patients with chronic diarrhoea [
]. We think that these discrepant results are due to the lack of immunofluorescent diagnostic criteria for EA. To look for EA is difficult. We would say that it is like distinguishing a black cat from a white one. It seems to be easy and indeed it is easy in the most obvious cases (Fig. 2), but sooner or later we will come across a grey cat, i.e. an intermediate pattern difficult to define as either black or white. So, on the basis of our experience, we recommend that only very bright patterns like those in Fig. 2 should be considered to be positive EA. Weak fluorescent staining of enterocytes, especially in the crypts, are on the other hand very common and totally non-specific. High EA titres do not correlate with histological severity and they might appear late during AIE natural history [
Impact of antibodies to infliximab on clinical outcomes and serum infliximab levels in patients with inflammatory bowel disease (IBD): a meta-analysis.
]. Moreover, the small group of pANCA-positive CrD patients have endoscopic, pathological and clinical features of left-sided colitis, thus indicating a predominant UC phenotype [
]. Among the subgroup of pANCA-positive IBD patients, the anti-bacterial flagellin CBir1 antibodies are expressed in 44% of CrD patients and only in 4% in UC cases, and this may help in discriminating UC from CrD [
], serum levels of the anti-glycan anti-Saccharomyces cerevisiae antibodies (ASCA) have a sensitivity of 56% and a specificity of 88% in discriminating CrD from UC. ASCA-positive CrD patients have a higher risk of disabling disease, including stricturing and penetrating behaviour, earlier onset and perianal disease, resulting in an increased need for surgery [
]. A recent study showed that faecal ASCA have the same sensitivity (52%) but a lower specificity (71%) compared to serum ASCA (87%) in a paediatric cohort of 83 CrD patients [
]. Apart from ASCA, five further serum anti-glycan antibodies, i.e. anti-chitobioside carbohydrate IgA (ACCA), anti-mannobioside carbohydrate IgG (AMCA), anti-laminaribioside IgG (ALCA), anti-laminarin carbohydrate antibodies (anti-L), and anti-chitin carbohydrate (anti-C), have been detected in the serum of IBD patients. It has been shown that serum levels of all the anti-glycan antibodies remain stable over six years of follow-up in most patients [
]. CrD patients positive for ASCA, AMCA and anti-L have a higher risk of complications, whereas anti-C is strongly associated with IBD-related surgery [
]. ALCA or ACCA were detected in 44% of a small cohort of ASCA-negative CrD patients from Israel, increasing the sensitivity and specificity up to 77% and 90%, respectively, in CrD patients positive for at least one out of the three antibodies, ALCA, ACCA and ASCA [
]. On the other hand, positivity for anti-Escherichia coli outer membrane protein C antibodies (anti-OmpC), which are frequently associated with a disabling disease course and surgery in CrD [
Antibodies to Escherichia coli outer membrane porin C in the absence of anti-Saccharomyces cerevisiae antibodies and anti-neutrophil cytoplasmic antibodies are an unreliable marker of Crohn disease and ulcerative colitis.
]. A recent study on twins with CrD showed a high degree of similarity in anti-OmpC and in anti-Pseudomonas fluorescens-associated peptide I2 antibodies (anti-I2) in discordant monozygotic twin pairs, but not in discordant dizygotic twin pairs, suggesting that both anti-OmpC and anti-I2 stand for a genetically determined loss of tolerance [
]. In addition, over six years before the diagnosis in 65% of CrD patients, at least one of ASCA IgA, ASCA IgG, anti-OmpC, anti-CBir1, and two other anti-flagellin antibodies, anti-Fla2 and anti-FlaX, is positive [
]. Similarly, a panel of serum antibodies, including pANCA, ASCA IgA, ASCA IgG, anti-OmpC, anti-CBir1, has been shown to predict the diagnosis of both CrD and UC in the previous four years [
]. Anti-microbial antibodies are clinically useful in predicting a disabling disease course in CrD, but not in discriminating between stricturing and non-stricturing phenotypes [
Due to its chimeric nature and consequent high immunogenicity, the monoclonal anti-TNF antibody infliximab may induce the development of ADA, generally within the first 12 months of therapy [
]. ADA, whose appearance may be delayed by concomitant administration of immunosuppressant drugs (i.e. thiopurines), induce and frequently precede the loss of response to infliximab [
], their titres inversely correlate with adalimumab concentration and, hence, high amounts of ADA are associated with disease activity due to their drug inactivation and clearance in CrD [
]. In the case of loss of response, a number of different strategies may be adopted on the basis of the presence or absence of ADA positivity and according to drug trough levels (Fig. 3) [
]. However, it has been proposed that adequate serum trough levels (3–7 μg/ml) do not necessarily reflect translated amounts of anti-TNF agents sufficient to neutralize all the TNF present in IBD mucosa [
]. Until now, no study has been conducted on ADA directed against golimumab in UC patients. Similarly, treatment with the humanized monoclonal antibody vedolizumab is associated with the development of ADA in 4% of patients [
]. According to our recent demonstration that increased amounts of matrix metalloprotease-3 and -12 in active IBD mucosa may cleave not only anti-TNF agents but also endogenous IgG at hinge region level, raised serum levels of matrix metalloproteinase-cleaved endogenous IgG and anti-hinge autoantibodies against neo-epitopes of cleaved IgG may be considered as predictors of poor response to anti-TNF therapy [
Proteolytic cleavage and loss of function of biologic agents that neutralize tumor necrosis factor in the mucosa of patients with inflammatory bowel disease.
Fig. 3Algorithm for optimization of anti-tumour necrosis factor (TNF) therapy in inflammatory bowel disease patients depending on serum anti-TNF antibody levels (therapeutic range: 3–7 μg/ml) and the presence or absence of anti-drug antibodies (ADA).
In conclusion, none of the aforementioned autoantibodies currently have enough accuracy to justify their use in day-to-day clinical practice, whereas ADA appear to be a useful tool in the clinical management of IBD patients exposed to biologic agents.
6. Autoimmune liver disease
Autoimmune serology has a central role in the diagnosis and classification of autoimmune liver disease (Table 2). However, there are a number of controversial issues as most autoantibodies characteristically associated with autoimmune liver disease lack specificity and pathogenic significance. Indirect immunofluorescence is the main technique for routine autoantibody testing [
]. It is based on the use of a freshly frozen rodent substrate that usually includes kidney, liver and stomach, a combination that allows the simultaneous detection of anti-nuclear antibodies (ANA), anti-smooth muscle antibodies (SMA), anti-liver kidney microsomal antibody (LKM) 1, anti-mitochondrial antibody (AMA), and anti-liver cytosol 1 (LC1), if LKM1 is absent. In adults, significant titres are ≥1:40 dilution by indirect immunofluorescence. In children, titres of 1:20 for ANA or SMA and 1:10 for LKM1 are supportive of the diagnosis of AIH when used in combination with other laboratory tests and clinical features suggestive of the disease. However, the indirect immunofluorescence technique has many drawbacks, as it is time consuming, requires an experienced observer, and is insufficiently standardized. Commercially available substrates are used in routine laboratory practice, but their quality varies. These substrates are treated with fixatives in order to lengthen their shelf life, but this also causes enhanced background staining, which might cause difficulties in the interpretation of fluorescence patterns and can explain the low reproducibility of the test. Methods other than indirect immunofluorescence, such as ELISA or immunoblotting, are gaining popularity. This shift has been supported by the introduction of assays based on recombinant/purified target antigens, such as cytochrome P 450 2D6-CYP2D6, formimino-transferase cyclo deaminase (FTCD), soluble liver antigen/liver pancreas, and filamentous actin-F-actin. However, the use of ELISA as the sole primary screening test is inappropriate because there is no useful combination of molecular specificities for a dependable detection of ANA and SMA, while the results are interchangeable with indirect immunofluorescence for those autoantibodies (AMA, LKM1 and LC1) whose target antigen has been identified at molecular level [
Autoantibody titres and specificity may vary during the course of the disease, and seronegative individuals at diagnosis may express the conventional autoantibodies later in the disease course [
], and do not need to be monitored regularly, unless a significant change in the clinical phenotype appears. On the contrary, in childhood autoantibody titres may be useful biomarkers of disease activity, and they can be used to monitor treatment response [
Organ and non-organ specific autoantibody titres and IgG levels as markers of disease activity: a longitudinal study in childhood autoimmune liver disease.
]. In particular, LC1 have been demonstrated to correlate well with disease activity showing a significant decrease in titre (>50%) or disappearance during remission and flare-up during relapse [
], are not disease-specific and show a wide range of heterogeneity in terms of antigenic specificity. The fluorescence pattern of ANA in AIH is usually homogeneous using HEp-2 cells, but a speckled pattern is rather frequent. ANA-positivity is found in 43% of type 1 AIH patients [
], and is associated with a variety of antigenic specificities including histones, double-stranded DNA (15%), chromatin and ribonucleoprotein complexes. However, no single pattern or combination is pathognomonic of AIH. Thus, the investigation of different ANA staining patterns has no clinical or diagnostic relevance in routine practice, and the use of HEp2 cells at AIH screening stage is not recommended [
Diagnostic accuracy of four different immunological methods for detection of anti-F-actin autoantibodies in type 1 autoimmune hepatitis and other liver-related disorders.
]. SMA react to several cytoskeletal elements including F-actin with a reported prevalence of anti-actin antibodies in 41% of patients. When kidney sections are used as a substrate for indirect immunofluorescence, SMAvg (vessel/glomerulus) and SMAvgt (vessel/glomerulus/tubule) patterns can be identified, which are frequently associated with AIH, though not pathognomonic. SMAgt correlate with F-actin antigenicity [
Diagnostic accuracy of four different immunological methods for detection of anti-F-actin autoantibodies in type 1 autoimmune hepatitis and other liver-related disorders.
]. However, indirect immunofluorescence remains superior to ELISA and provides the best accuracy. Indeed, actin is not the only target antigen of AIH-specific SMA reactivity, and thus ELISA can miss the diagnosis in about 20% of cases [
]. ANA and SMA reactivity frequently coexist in the same serum, thus improving the strength of the diagnosis.
LKM1 and/or LC1 are the serologic markers of type 2 AIH. The two antibodies often coexist in the same serum, but in some cases LC1 is present alone and is the only marker for the diagnosis of type 2 AIH. The major target autoantigen of LKM1 has been clearly identified as the cytochrome P4502D6 (CYP2D6) and the FTCD for LC1. However, neither LKM1 nor LC1 are highly disease-specific, as they have been described in a small proportion (5–10%) of adult and paediatric patients with chronic HCV infection [
Cytochrome P4502D6 (193–212): a new immunodominant epitope and target of virus/self cross-reactivity in liver kidney microsomal antibody type-1 positive liver disease.
]. Anti-soluble liver antigen/liver-pancreas antibodies (SLA/LP) are the only disease-specific biomarker. The target antigen has been identified as a synthase (S) converting O-phosphoseryl-tRNA (Sep) to selenocysteinyl-tRNA (Sec), named as SepSecS [
], but sometimes it is the only detectable autoantibody reactivity. SLA/LP has never been described in association with LKM1 and LC1 in type 2 AIH. The positivity of pANCA, originally considered specific for PSC and IBD, is also frequently present in patients with type 1 AIH [
]. A recent meta-analysis indicated that ANA have moderate sensitivity and specificity, SMA have moderate sensitivity and high specificity, and SLA/LP have low sensitivity and high specificity [
Meta-analysis: diagnostic accuracy of antinuclear antibodies, smooth muscle antibodies and antibodies to soluble liver antigen/liver pancreas in autoimmune hepatitis.
AMA at titres higher than 1:40 have a high sensitivity and specificity for the diagnosis of PBC, being present in 90–95% of patient sera. AMA are in fact one of the diagnostic criteria of PBC along with elevated alkaline phosphatase and a compatible liver histology [
Meta-analysis: diagnostic accuracy of antinuclear antibodies, smooth muscle antibodies and antibodies to soluble liver antigen/liver pancreas in autoimmune hepatitis.
]. AMA are typically detected by indirect immunofluorescence, characterized by the staining of all three substrates, namely smaller distal renal tubules, gastric parietal cells and liver cell cytoplasm. AMA target the 2-oxoacid dehydrogenase complex family. The major epitope (AMA-M2) is located on the subunit of the pyruvate dehydrogenase complex-E2, but AMA also react with the other two components of the pyruvate dehydrogenase complex. Immunoblotting and ELISA tests are now available, and these assays have an increased sensitivity and specificity (greater than 95%) [
]. The role of AMA in the pathogenesis of PBC is under debate because of the accessibility of the autoantibody to an antigen located in the inner membrane of mitochondria. Recent evidence indicates that AMA recognize pyruvate dehydrogenase complex-E2 in apoptotic bodies resulting in a complex that stimulates innate immune systems in genetically susceptible individuals [
]. AMA reactivity can be detected decades before the clinical onset of PBC; however, 76% of asymptomatic patients with incidental AMA serum reactivity eventually develop PBC over more than 10 years of observation [
] in patients with the classic phenotype of AIH without any other evidence of PBC, and may hint at co-existent or underlying PBC. Nonetheless, these patients should be classified and treated according to their clinical phenotype.
Non-specific ANA are found in at least 30% of PBC sera [
] without an apparent correlation with diagnosis or disease phenotype. However, some ANA reactivity with rim-like/membranous or multiple nuclear dot immunofluorescence patterns are highly specific to PBC. The identified targets are the nuclear body 100 kDa (sp100), promyelocitic leukaemia and small ubiquitin-like modifiers corresponding to the multiple nuclear dot-ANA, and proteins within the nuclear pore complex, including the 210 kDa glycoprotein (gp210) and the 62 kDa nucleoprotein corresponding to the rim-like/membranous pattern [
Unlike AIH and PBC, autoantibody testing is not included in the diagnostic work-up of PSC. The most frequent autoantibody found in patients with PSC is pANCA, which is routinely detected by an indirect immunofluorescence assay. The target antigen is located in the nuclear membrane [
Differentiation of antineutrophil nuclear antibodies in inflammatory bowel and autoimmune liver disease from anti-neutrophil cytoplasmic antibodies (p-ANCA) using immunofluorescence microscopy.
], but the antigens targeted by pANCA in PSC are still unknown. Several cytoplasmic proteins have been proposed, including lactoferrin, myeloperoxidase, cathepsin G, proteinase 3 and catalase [
] because the prevalence can be found at equal rates in AIH and PBC. No association links pANCA to the genetic susceptibility of PBC in terms of a particular HLA haplotype [
] has reported a higher prevalence of pANCA in patients with PSC and IBD than in patients without IBD. The prognostic role of pANCA in PSC has also been investigated, but, even though studies seem to suggest an association between pANCA and end-stage liver disease [
], most studies reported no differences in pANCA-positivity between early and advanced PSC, and a significant correlation between titres and disease activity has not been clearly demonstrated [
A 2-year follow-up study of anti-neutrophil antibody in primary sclerosing cholangitis: relationship to clinical activity, liver biochemistry and ursodeoxycholic acid treatment.
]. None of the autoantibodies described in PSC have sufficient specificity and sensitivity to be used for screening or diagnosis.
In conclusion, although autoimmune liver disease encompasses a wide – and sometimes overlapping – range of immune-mediated disorders, AMA and SLA/LP are the only disease-specific antibodies for these conditions. Therefore, in the case of overlapping autoimmune characteristics, making an appropriate diagnosis might still be tricky and it should rely on clinical, serological, radiological and histological features.
7. Autoimmune pancreatitis
AIP is a fibro-inflammatory condition involving the pancreas and, sometimes, extra-pancreatic organs. There are two distinct subtypes of the disease (type 1 and type 2), which clearly differ in histological, clinical and epidemiological features. Currently, the International Consensus Diagnostic Criteria are widely accepted for the diagnosis of AIP subtypes, and they are based on five cardinal features: (i) pancreatic parenchymal and ductal imaging, (ii) serology, (iii) other organ involvement, (iv) pancreatic histology, and (v) response to steroid therapy [
]. In the absence of a definitive histological diagnosis, which generally allows the distinction between type 1 and type 2 AIP, a combination of several criteria is needed for the diagnosis of type 1 AIP. Serological abnormalities and other organ involvement are consistent with type 1 AIP [
]. Since type 1 accounts for 80–90% of AIP and type 2 for only 10–20%, increased serum levels of IgG4 are observed only in up to 70% of AIP patients. Furthermore, IgG4 elevation may be observed in up to 10% of patients suffering from other diseases, such as pancreatic cancer, cholangiocarcinoma and PSC [
]. Indeed, the specificity of serum IgG4 in the diagnosis of AIP is higher if the elevation is more than twice the upper limit of normal, and it increases proportionally with the serum IgG4 level. Considering the low prevalence of the disease (4.6/100.000) [
], serum IgG4 should be managed carefully in clinical practice, in order to avoid misdiagnoses, and should only be considered as a part of the diagnostic algorithm [
Many studies have proposed other biomarkers for the diagnosis of AIP, but none have reported satisfactory specificity and sensitivity levels, and their use is still limited to research protocols. Up to 40% of AIP patients have ANA [
] were identified in 55%, 75% and 33% of patients, respectively. Other autoantibodies described in AIP patients are against trypsinogens PRSS1 and PRSS2 [
In conclusion, the only useful biomarker in clinical practice for AIP is serum IgG4. Considering its low specificity and sensitivity, other criteria are needed for the diagnosis of this disorder. New serological markers are expected in order to improve, alone or in combination with serum IgG4, the diagnosis of a disease that still remains difficult to recognize.
8. Concluding remarks
Based on recent advances in most fields of medicine we can expect increasing progress in the future in the development of more reliable serum antibody biomarkers for diagnosing immune-mediated gastrointestinal and liver diseases. These new tests should make it possible to predict a patient’s risk of developing a disease, facilitate early diagnosis, promote case-finding strategies, allow assessment of patient prognosis, and prove the efficacy of therapeutic strategies. This is particularly the case in those conditions, such as AIE, IBD, AIH, PSC and AIP, in which serum antibodies are of little use because of their poor diagnostic accuracy.
Key messages
–
Case-finding is a good strategy in coeliac disease and autoimmune atrophic gastritis based on the measurement of their specific serum antibodies – anti-tissue transglutaminase and anti-endomysial antibodies for the former and anti-gastric parietal cell antibodies for the latter condition – followed by histologic confirmation on duodenal and gastric biopsy, respectively.
–
Anti-enterocyte antibodies are virtually 100% sensitive, but the lack of immunofluorescent diagnostic criteria may decrease their accuracy for the diagnosis of autoimmune enteropathy.
–
Optimization of anti-tumour necrosis factor therapy in patients with inflammatory bowel disease relies on serum drug levels and the presence or absence of anti-drug antibodies.
–
Unlike autoimmune hepatitis and primary biliary cholangitis, the diagnostic work-up of primary sclerosing cholangitis does not encompass autoantibody testing, and only anti-mitochondrial antibodies and anti-soluble liver antigen/liver-pancreas antibodies are disease-specific (for primary biliary cholangitis and autoimmune hepatitis, respectively).
–
The only available marker for autoimmune pancreatitis is serum IgG4, and considering the low prevalence of the disease and the difficult differential diagnosis with cancer, this marker should be managed carefully in clinical practice in order to avoid misdiagnoses.
Conflict of interest
None declared.
References
Di Sabatino A.
Lenti M.V.
Giuffrida P.
et al.
New insights into immune mechanisms underlying autoimmune diseases of the gastrointestinal tract.
Gastric parietal cell antibodies, Helicobacter pylori infection, and chronic atrophic gastritis: evidence from a large population-based study in Germany.
Cancer Epidemiol Biomarkers Prev.2013; 22: 821-826
Impact of antibodies to infliximab on clinical outcomes and serum infliximab levels in patients with inflammatory bowel disease (IBD): a meta-analysis.
Antibodies to Escherichia coli outer membrane porin C in the absence of anti-Saccharomyces cerevisiae antibodies and anti-neutrophil cytoplasmic antibodies are an unreliable marker of Crohn disease and ulcerative colitis.
Proteolytic cleavage and loss of function of biologic agents that neutralize tumor necrosis factor in the mucosa of patients with inflammatory bowel disease.
Organ and non-organ specific autoantibody titres and IgG levels as markers of disease activity: a longitudinal study in childhood autoimmune liver disease.
Diagnostic accuracy of four different immunological methods for detection of anti-F-actin autoantibodies in type 1 autoimmune hepatitis and other liver-related disorders.
Cytochrome P4502D6 (193–212): a new immunodominant epitope and target of virus/self cross-reactivity in liver kidney microsomal antibody type-1 positive liver disease.
Meta-analysis: diagnostic accuracy of antinuclear antibodies, smooth muscle antibodies and antibodies to soluble liver antigen/liver pancreas in autoimmune hepatitis.
Differentiation of antineutrophil nuclear antibodies in inflammatory bowel and autoimmune liver disease from anti-neutrophil cytoplasmic antibodies (p-ANCA) using immunofluorescence microscopy.
A 2-year follow-up study of anti-neutrophil antibody in primary sclerosing cholangitis: relationship to clinical activity, liver biochemistry and ursodeoxycholic acid treatment.
30940-4&id=gr1.jpg){kind=link}
30940-4&id=gr2.jpg){kind=link}
30940-4&id=gr3.jpg){kind=link}