Article Text

ANCAs away
  1. Institute of Liver Studies
  2. King’s College Hospital
  3. Denmark Hill, London SE5 9RS, UK
  4. email: ian.mcfarlane{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Significant titres of circulating antibodies reacting with a wide range of normal tissue components occur with variable frequency in several liver disorders. Those that are of particular relevance to liver disease include the antinuclear (ANA), smooth muscle (SMA), liver–kidney microsomal (LKM), antimitochondrial (AMA), and antineutrophil cytoplasmic (ANCA) antibodies. In some conditions, such as autoimmune hepatitis (AIH) and primary biliary cirrhosis (PBC), they form part of the diagnostic criteria.1 ,2 Their presence in others, such as chronic viral hepatitis, may serve to confuse. These autoantibodies are routinely detected by indirect immunofluorescence on sections of appropriate tissues or isolated cells and an experienced eye is required for interpretation of the different patterns of immunofluorescent staining. It has long been recognised, however, that antibodies giving apparently similar staining patterns often react with different antigens in the substrates used for their detection. In order to improve the sensitivity and specificity of these antibodies for different liver disorders, and to provide more objective methods for their detection, considerable efforts have been made in recent years to identify the specific antigens with which they react, with some success. Thus, the cytochrome P450 isoform IID6 has been identified as the target of LKM antibodies that are particularly associated with so called type 2 AIH,3 whereas the AMA that are most specific for PBC are now known to react with the E2 components of the mitochondrial 2–oxoacid dehydrogenase complexes.2

ANCA, which have been extensively studied in patients with segmental necrotising glomerulonephritis, Wegener’s granulomatosis and other vasculitides, are classified into two major subtypes according to whether they give cytoplasmic (cANCA) or perinuclear (pANCA) patterns of immunofluorescent staining on ethanol fixed neutrophils.4 ,5 cANCA reacting with proteinase 3 are particularly associated with Wegener’s granulomatosis and some classic pANCA are specific for myeloperoxidase but both subtypes are notably heterogeneous and a number of other target autoantigens have been identified, including azurocidin, lactoferrin, and cathepsin G.4 ,5 Additionally, some pANCA (“atypical” or xANCA) give staining patterns that are subtly different from those of classic pANCA, implying reactivities with yet other antigens.5

The relevance of pANCA to liver disease relates to the observation that these antibodies occur in more than 80% of patients with primary sclerosing cholangitis (PSC).6 For a time, they were thought to be a potential diagnostic marker of this disease,7 but recent studies have indicated that high titres of pANCA also occur with similar frequency in AIH.8These two conditions can usually be differentiated on the basis of histological and cholangiographic findings but they share several autoimmune features which can complicate the diagnosis. If the antigenic specificities of pANCA in PSC and AIH differ, and can be shown to be distinct from pANCA in other conditions, specific tests for these reactivities would be a useful addition to the diagnostic repertoire. Two very recent studies suggest that some progress is being made in this regard. The first indicates that atypical pANCA associated with ulcerative colitis, PSC and AIH react with nuclear lamins A, C and B1, and with the lamin B receptor, and that these reactivities are never observed in systemic vasculitis.9

The second report, which appears in this issue (see page 867), provides persuasive evidence that nearly all patients with AIH with classic pANCA have antibodies that react with the non-histone chromosomal proteins HMG1 or HMG2, or both. How these antibodies to HMG relate to pANCA is, however, uncertain. Absorbing the sera with purified HMG only partially abrogated pANCA staining, perhaps because most of the sera also contained antibodies reacting with two other known pANCA targets: cathepsin G and lactoferrin. Conversely, and somewhat surprisingly, 64% of PBC sera that were negative for classic pANCA by immunofluorescence (despite having antibodies against cathepsin G or lactoferrin) were positive for anti-HMG. Unfortunately, the authors did not include patients with PSC in their main study group although in the discussion they cite preliminary data indicating that antibodies to HMG may occur much less frequently (two of eight) in pANCA positive patients with PSC. Thus, neither of these two reports provides definitive evidence identifying subspecificities of pANCA that may be useful in the differential diagnosis of hepatobiliary diseases but both provide sufficiently suggestive data to indicate that further studies along these lines in much larger groups of patients with a wider range of liver disorders are worthwhile.


Linked Articles