Human leucocyte antigen‐B27 and ankylosing spondylitis

Human leucocyte antigen (HLA)-B27 (B27) is the major gene determining susceptibility to ankylosing spondylitis (AS) and the seronegative arthropathies, including reactive arthritis, psoriatic spondyloarthritis and colitic arthritis. B27 is also involved in susceptibility to anterior uveitis and is thought to be partially protective against HIV. AS occurs in 0.4% of western Caucasians,1 and it has been estimated that there are 22 000 Australian adults with AS, of whom only approximately one-third have been diagnosed.2 The prevalence of AS is generally proportional to the prevalence of B27 in different world populations. Thus AS is common in Inuit and some North American Indian populations where B27 prevalence can exceed 20%, but is rare in most Africans and Australian Aborigines, where B27 is very uncommon (<0.5% prevalence). In British Caucasian populations, ∼8% of the population carries B27, but only ∼5% of carriers develop the disease. There are now more than 40 different allelic variants of HLA-B27 reported, all of which are ancestrally related to B*2705. This variant is found in 95% of British Caucasians. There is convincing evidence that, in most ethnic groups, the subtype of B27 carried has no clinical relevance, with each common subtype having a similar risk of developing AS. Certainly this is the case for the main Caucasian subtypes, B*2705 and B*2702, which are equally strongly associated with AS.3 Two subtypes, HLA-B*2706 and B*2709, have been shown to have reduced penetrance for AS.4, 5 No cases of AS have yet been documented that carry B*2709, whereas the B*2706 subtype is incompletely protective and a small number of cases have been reported carrying this allele. B*2709 is common in Sardinia, but is very rare elsewhere. B*2706 is typically found in people of Asian descent and is particularly common in Malay ethnic groups compared with Chinese.5 In this issue of the Journal, Hou and Lai present data suggesting that B*2704 may be more strongly associated with AS than other subtypes present in Taiwanese Chinese, and that AS can occur in carriers of B*2706, confirming that the protective effect of this B27-subtype is not absolute.6 The marked effects that minor allelic variations in B27 have on penetrance have provided useful clues in research into the mechanism underlying the association of B27 with AS, which remains unexplained. So what is the role of B27-typing in clinical settings? In established AS, B27-typing can be useful to assist in estimating the likelihood of disease occurrence in relatives. Roughly 10% of B27-positive first-degree relatives of AS cases will also develop AS, whereas the chance of occurrence in B27-negative family members is remote. B27-negative cases have later age of disease onset (by approximately 10 years) and are less often affected with anterior uveitis, but otherwise their clinical course is the same as B27-positive cases.7 The main clinical role for B27-testing is to assist with diagnosis, but its use in this setting causes much confusion. Carriage of B27 alone does not diagnose AS – indeed 95% of carriers do not have a B27-related disease. However, B27 is associated with AS with an odds ratio of >100, so carriers are at far greater risk of developing the condition. For clinicians, the most relevant statistic is the probability that someone has AS, given the B27 result and the pretest probability that a patient has the disease (Fig. 1). Among patients presenting to a general practitioner with low back pain, the probability of AS is low (∼5%). In this setting, a B27-positive case will have ∼30% likelihood of having AS, whereas if the B27 test is negative, then the chance of disease is remote (<1%). If the pretest probability of AS is 50%, then a positive B27 test would increase the likelihood of AS to 90% and a negative test reduce the likelihood of the disease to 10%. Complex algorithms have been developed to determine the pretest probability of AS depending on the clinical circumstance, but these have not been tested clinically and are based on univariate analyses and therefore do not take into account interactions and correlations between clinical variables.8 Whether they are any better than a clinician’s best guess as to their certainty regarding a patient’s diagnosis is therefore debatable. Probability of ankylosing spondylitis (AS) given the likelihood of the disease before B27-testing and whether the individual carries B27. , probability of AS if B27 test negative; , probability of AS if B27 test positive. There is a great need to develop and validate diagnostic algorithms for AS, particularly in early disease. Average diagnostic delay in AS is currently 8–11 years,9 and it takes on average >9 years for diagnostic radiographic changes to develop on average in AS from symptom onset.10 Thus X-rays are of little use in diagnosis early in the disease course. Although magnetic resonance imaging (MRI) scanning is sensitive in early disease, its specificity for AS has not yet been established and diagnostic criteria have not yet been developed that incorporate MRI. This is a major stumbling block in access to tumour necrosis factor-antagonist therapy, which has revolutionized management of AS, but access to which is currently restricted to those with radiographic sacroiliitis. As the patients with the most to gain from disease suppression are those with the least joint damage, improved criteria for use in early AS are much needed. Similarly, the availability of effective treatments provides yet another reason why clinicians need to be more aware of this common disease to reduce the high proportion of undiagnosed cases and the lengthy average diagnostic delays currently experienced by people with AS.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.