Abstract

The vasculitides are a set of related uncommon but not rare disorders characterised by inflammation of blood vessel walls leading to damage to tissue or end organ. Classification of these disorders has been challenging for many years. The 1990 American College of Rheumatology criteria for classifying primary vasculitides was based on a large dataset comparing the clinical and laboratory features of different patients with a variety of forms of vasculitis. There was substantial variation in the sensitivity (71-95.3%) and specificity (78.7-99.7%). Unfortunately these criteria did not apply to the condition microscopic polyangiitis (MPA) which had been incorrectly included in the group of patients with polyarteritis nodosa (PAN). The distinction between these two conditions is very important because they are very distinct in their pathogenesis, clinical pattern and response to therapy. For example, we know that many patients with MPA will respond to rituximab whereas this would not be an appropriate treatment for PAN. We know more about the epidemiology and pathogenesis of these diseases; PAN (when it is not associated with Hepatitis B) is actually a very rare disorder compared to MPA. This perhaps raises the issues between disease definitions and classification criteria. Disease definition have been developed by the Chapel Hill consensus group to try and separate out different disease entities and to provide clear descriptions of MPA as separate from PAN as well as for other types of primary systemic vasculitis. However, the confusion continues between classification criteria and diagnostic criteria. In many instances existing classification criteria are adopted in a wide spread manner and applied by a clinician for diagnosis which is not an appropriate use of classification criteria. Diagnostic criteria would require appropriate non-disease controls to compare against, whereas classification criteria are usually developed without any controls. There has been an evolution in diagnostic techniques particularly in imaging and a better understanding of the pathophysiology of vasculitis. Applying what we know now should result in improvements in classification and help to inform the development of diagnostic criteria. We are undertaking a multicentre, multinational study using data driven methods to develop both a revised classification system for vasculitis and by including a control group, we can also develop diagnostic criteria. The study will be described in detail (Diagnosis and classification of vasculitis, DCVAS) we plan to include over 2000 patients with primary systemic vasculitis and over 1500 patients with conditions that may mimic vasculitis. Currently we have recruited over 900 patients from over 40 centres within Europe, North and South America, Australasia and Japan. There will probably never be a “gold” standard for classification or diagnostic criteria but we may achieve “close to gold” standards by using a physician based diagnostic model with predictive values for sensitivity and specificity and likelihood ratio of each particular disease and this should lead to useful and practical application for both researchers and clinicians managing patients with systemic vasculitis.