Kel140 1549.1554

Prevalence and management of rheumatoid arthritis in thegeneral population of Greece—the ESORDIG study A. Andrianakos1,2, P. Trontzas3, F. Christoyannis1, E. Kaskani4, Z. Nikolia5,E. Tavaniotou1, A. Georgountzos3 and P. Krachtis1 for the ESORDIG study groupy Objective. To assess the prevalence and management of rheumatoid arthritis (RA) in the general adult population of Greece.
Methods. This cross-sectional study was conducted on the total adult population (!19 yrs old) of seven communities(8547 subjects), and on 2100 out of 5686 randomly selected subjects in two additional communities. The study, based ona standardized questionnaire and clinical evaluation and laboratory investigation when necessary, was carried out byrheumatologists who visited the target population at their homes. Diagnosis of RA was based on the American College ofRheumatology (ACR) 1987 criteria.
Results. A total of 8740 subjects participated (response rate 82.1%). RA was diagnosed in 59 individuals. The prevalence of RAwas 0.68% (95% CI 0.51–0.85); it was significantly higher in females than males (P < 0.0005), and increased significantlywith age up to and including the 50–59-yr-old group (P < 0.002), and then decreased slightly. On their first medical visit, 19%(95% CI 9.7–30.9) of the RA patients had consulted a rheumatologist, while during the first year after disease onset, 61%(95% CI 48.6–73.4) had done so. Early consultation with a rheumatologist and disease-modifying anti-rheumatic drug(DMARD) combination therapy were negatively associated with ACR functional classes II–IV [adjusted odds ratios 0.18(95% CI 0.04–0.85) and 0.17 (95% CI 0.04–0.72), respectively].
Conclusions. The prevalence of RA in the general adult population of Greece is similar to that in many other Europeancountries; early consultation with a rheumatologist and DMARD combination therapy are associated with a better RAoutcome.
KEY WORDS: Rheumatoid arthritis, Prevalence, Epidemiology, Management, Greece.
Rheumatoid arthritis (RA) is a chronic and deforming inflam- disability. Epidemiological studies have shown that the prevalence Details on the ESORDIG study population, subject recruitment of RA varies broadly from 0.2 to 1.0% in various European, and evaluation, as well as on quality control have been reported North American, Asian and Australian populations [1]. Most previously [14]. The ESORDIG study was conducted from March studies in European countries have suggested a prevalence in adult 1966 to April 1999 on the total adult population (aged !19 yrs populations ranging from 0.5 to 1.0% [1–7]. However, some old) of two urban, one suburban and four rural areas located in studies, especially those from southern European countries, northern, central and southern mainland Greece (8547 subjects), including Greece, have shown a lower prevalence (0.18–0.34%), as well as on 2100 out of 5686 randomly selected adult subjects in which raises important questions about the possible involvement one additional rural and one suburban community. In the latter of different environmental and/or genetic factors in the aetiology areas, every second and third household from a randomly chosen of RA among various European populations [8–10]. Few starting point, respectively, was selected (systematic sampling) population-based studies have assessed the care of RA patients (Fig. 1); this was for practical reasons since there were only two [11–13] and data on the association between care and the outcome investigators available for the suburban and one for the rural area.
of RA in the general population are limited.
Sixteen rheumatologists conducted the study by visiting the target This part of the ESORDIG (epidemiological study of the population at their homes. Each visit involved an interview with rheumatic diseases in Greece) study aimed at assessing the each participant that was based on a standardized questionnaire prevalence and management of RA in the general adult aimed at obtaining a variety of information on socio-demographic characteristics, medical history, and on a specific standardized 1Rheumatic Disease Epidemiology Section, Hellenic Foundation for Rheumatological Research, 2Third Department of Internal Medicine, AthensUniversity Medical School, Sotiria Hospital, 3Rheumatology Department, 3rd IKA Hospital, 4IKA Health Center, Halandri and 5DEH Health Center,Athens, Greece.
Received 18 November 2005; accepted 16 December 2005.
yIn addition to the authors, the following physicians are members of the ESORDIG study group: P. Dantis, D. Karamitsos, G. Kaziolas, L. Kontelis, Correspondence to: A. Andrianakos, Hellenic Foundation for Rheumatological Research, 8 Rodon Street, Kantza Pallini Attikis, 153 51 Athens, ß The Author 2006. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected] Eligible adult population (19 years)
n = 14233
Total adult populationin 7 areas
Selected samplein 2 areas
[Suburban: 1486 out of4458 (ratio 1:3) n = 8547
Rural: 614out of 1228 (ratio 1:2)] n = 2100
Final target adult population
n = 10647
Participants
(Interviewed and evaluated by rheumatologists attheir homes) n = 8740
Participation rate: 82.1%)
FIG. 1. Flow chart showing the ESORDIG study design.
questionnaire aimed at revealing all subjects suffering from RA.
This specific questionnaire was analogous to that used byMacGregor et al. [2] and consisted of the following three All analyses were conducted using SPSS v.12.0 for Windows. The questions: have you ever had (i) any joint pain, not due to chi-square test was used to compare prevalence and percentages, trauma, lasting at least six continuous weeks? (ii) any joint while the comparison of mean values was by Student’s t-test.
Values of P < 0.05 were considered significant; 95% confidence swelling lasting at least six continuous weeks? (iii) morning intervals (CIs) were given where relevant. A logistic regression stiffness in any joint, lasting at least 1 hour before maximal model was used for assessing the association of RA with certain improvement? The sensitivity of this questionnaire to detect cases factors such as sex, age, marital status, body mass index (BMI), of RA was shown to be 100% in a pilot study of 45 patients with cigarette smoking (pack-yrs), alcohol consumption, level of known RA, performed prior to the start of the ESORDIG study.
education, occupation, socioeconomic status and residence in All subjects who responded positively to any of the three questions urban, suburban or rural areas. Concerning BMI, cut-off points of of this specific questionnaire were subsequently evaluated by !30 kg/m2 for obesity and <30 kg/m2 for non-obesity were used the rheumatologists conducting the study (medical history, [18]. The level of education was defined as low or high on the basis clinical examination, assessment of available laboratory and of school attendance up to 9 and >9 yrs, respectively. Multiple imaging findings), during the same home visit. When necessary, logistic regression analysis was also applied for assessing the appropriate X-ray investigation and/or other requisite laboratory association of ACR functional classes II–IV with certain factors tests were performed on the following days, and the findings such as sex, age, residence, BMI, disease duration, disease were assessed by the rheumatologists during a second home visit, remission or not, presence of rheumatoid factor or not, early or in order to reach a definite diagnosis. The diagnosis of RA was late consultation with a rheumatologist and disease-modifying made on the basis of the American College of Rheumatology anti-rheumatic drug (DMARD) combination therapy.
consequences of RA were assessed using the respective ACRcriteria [16, 17].
The effect of non-selection and random selection of suburban and rural populations on the study results was tested in a logisticregression model in which the dependent variable was the Of the final target adult population of 10 647 subjects, 8740 diagnosis of RA and the independent variables were the participated in the study (participation rate 82.1%). Among the selected/non-selected populations. As previously described [14], participants, 4269 (49%) were men and 4471 (51%) were women, data were obtained from a random sample of non-responders on while 31% were residents in urban, 34% in suburban and 35% socio-demographic characteristics, past medical history, previous in rural areas; the age range was 19–99 yrs, mean 47 yrs (S.D. 17.7).
rheumatic disease diagnosis including RA, and the reasons for As reported previously [14], using Pearson correlation coefficients, we found significant similarities in terms of age and sexdistribution between the study participants, the total targetadult population and the total adult population of Greece, even when the data were analysed separately for urban, suburban and The study was conducted according to the declarations of rural populations. Logistic regression showed no effect of non- Helsinki and written informed consent was obtained from selection and random selection of suburban and rural populations all the study participants. The protocol was approved by the on the study results. Moreover, no significant difference was appropriate committees of the Ministry of Health and the Central found between non-responders and responders in terms of age, Union of Municipalities and Communities of Greece.
sex and prevalence of rheumatic symptoms or disease. The reasons Prevalence and management of RA in Greece TABLE 1. Demographic and clinical variables of the RA patients Values are percentages (95% CIs) unless otherwise stated; NS ¼ not significant.
for non-participation were unrelated to the presence or not ofrheumatic disease.
Of the 8740 participants, 59 were diagnosed as having had RA(Table 1). Thus, the age- and sex-adjusted prevalence of RA in thetotal target adult population was 0.67% (95% CI 0.54–0.80),while the prevalence of RA among the study participants was0.68% (95% CI 0.51–0.85). The prevalence of RA was signifi-cantly higher among females (1.0%, 95% CI 0.71–1.29) comparedwith males (0.3%, 95% CI 0.14–0.46) in the study participants(P < 0.0005), with a ratio of 3.3:1. The prevalence of RA increasedsignificantly with age up to and including the 50–59-yr-old group(P < 0.002), and then decreased slightly but non-significantly inthe last two age groups (P ¼ 0.44) (Fig. 2). There was nosignificant difference in the prevalence of RA among the urban,suburban and rural populations, nor between the selected and FIG. 2. Prevalence of RA by age group.
non-selected populations, nor even between the studied northern,central and southern areas of the country.
Logistic regression analysis showed that among the many TABLE 2. Medical specialties first visited by the 59 RA patients factors included in the model, only female sex and age !40 yrswere significantly associated with RA [adjusted odds ratios 3.7 (95% CI 2.0–6.9), P < 0.0005, and 6.1 (95% CI 2.8–13.4), Two of the 59 RA patients (3%, 95% CI 0.4–11.7) had not beenseen by a physician prior to the study and were diagnosed by theinvestigators. Although the other 57 RA patients had soughtmedical assistance for their symptoms, on their first medical visit administered in 21 patients and the most commonly used combi- only 11 patients (19%, 95% CI 9.7–30.9) had consulted a nations were hydroxychloroquine þ sulfasalazine þ methotrexate rheumatologist and the remaining 46 (78%, 95% CI 67.4–88.6) in six patients (29%), hydroxychloroquine þ methotrexate in five had seen physicians of other specialties (Table 2). However, most patients (24%), and methotrexate þ ciclosporin in four patients of the RA patients were seen by rheumatologists at subsequent (19%). Leflunomide and biological therapy were not available in medical visits, and remained under their care: 36 of the RA Greece at the time the study was conducted.
patients (61%, 95% CI 48.6–73.4) had consulted a rheumatologistduring the first year of the course of the disease (group I) and18 (30%, 95% CI 18.3–41.7) after the first year of the course of the disease (group II), while five patients (9%, 95% CI 2.8–18.7) hadnever seen a rheumatologist. Table 3 shows the demographic and In this part of the ESORDIG study, the prevalence and clinical variables for the RA patients in groups I and II. Multiple management of RA were assessed in urban, suburban and logistic regression analysis showed a significant negative associa- rural general adult populations of Greece. Among the study tion of an early consultation with a rheumatologist and of participants the RA prevalence was 0.68%; this was significantly DMARD combination therapy with ACR functional classes II–IV higher among women than men, and increased significantly [adjusted odds ratios 0.18 (95% CI 0.04–0.85), P < 0.031, and 0.17 with age up to and including the 50–59-yr-old group, and then (95% CI 0.04–0.72), P < 0.016, respectively].
decreased slightly. An early consultation with a rheumatologist Prior to being seen by a rheumatologist, 25 RA patients had and a DMARD combination therapy were negatively associated been treated at different times by at least two non-rheumatologist with ACR functional classes II–IV.
physicians. Comparative data on the diagnosis and treatment of Our estimate of RA prevalence is comparable with that found the RA patients by rheumatologists, orthopaedists and internists in other population-based studies in European Caucasians, which are shown in Table 4. The five most commonly prescribed used the same classification criteria [15]: 0.8% in Finland [3] and DMARDs in 52 patients were: methotrexate (81%), hydroxy- Manchester and Norfolk, UK [2, 19], 0.62% in Brittany, France chloroquine (46%), gold salts (37%), sulfasalazine (23%) and [4], and $0.5% in Sweden [5], Oslo, Norway [6], and Spain [7].
In a few studies from southern Europe [8–10], including Greece, TABLE 3. Demographic and clinical variables of the RA patients by early or late consultation with a rheumatologist* Values are percentages (95% CIs) unless otherwise stated; NS ¼ not significant.
*Group I: patients consulted a rheumatologist during the first year of the disease course; Group II: patients consulted a rheumatologist after the first TABLE 4. Diagnosis and treatment of the RA patients by rheumatologists and non-rheumatologists* NSAIDs: non-steroidal anti-inflammatory drugs; DMARD: disease-modifying antirheumatic drug.
*Prior to their visit to a rheumatologist, 32 patients had been treated by orthopaedists and 30 by internists. In addition, two patients had been continuously followed up by orthopaedists.
**Analgesics, local or intra-articular injections of corticosteroids.
and in a recent report from France [20], a lower prevalence of RA in European, North American, Asian and Australian epidemio- (0.18–0.34%) has been reported. Although this low prevalence logical population studies, with a female to male ratio varying could be related to a variation in genetic and/or environmental in the range of 2–5.6:1 [1, 4, 6–10, 20, 22, 23]. In accordance with risk factors in these areas, it seems more possible, however, that previous studies [1, 4, 7, 8, 22], RA prevalence increased with age reaching a peak in the 50–59-yr age-group. The slight decline of instance, the study in the Ioannina district of northwest Greece RA prevalence at older ages, we found, has also been reported in was based on RA cases diagnosed in two hospitals and private previous studies [4, 7, 8]; this could be attributed to an increased rheumatologists’ offices [8]. Thus, an underestimation of the RA mortality rate in RA patients at these ages [26, 27]. The residential prevalence seems quite possible, since patients with severe RA area did not affect the prevalence of RA in our study. However, could have moved and sought healthcare in other cities outside some studies have suggested that rural residence may be northwest Greece, while mild cases in the community could have associated with a lower prevalence [7, 28]. Whether a variation remained undiagnosed or they could have been under the care of in environmental or socioeconomic factors could be responsible other medical specialties [21]. Indeed, it was shown in the present for these differences is unknown, although no association between study that 9% of the RA patients had never been seen by a socioeconomic status and RA was found in our study.
rheumatologist, while during the first year of their disease course Prior to the present study, most of the RA patients had been only 61% of the patients had consulted rheumatologists. An treated by a rheumatologist. However, on their first medical visit, underestimation is also possible in the Belgrade study [9]; 18% of only a small percentage of RA patients (19%) had consulted the subjects with rheumatic complaints refused to undergo clinical rheumatologists, while within the first year after disease onset, evaluation, while patients with RA in remission were apparently 61% had visited rheumatologists; the latter finding is comparable not included in the prevalence estimation, since the questionnaire with that of a recent study from Germany [12]. This delay in used focused on symptoms during the 3 months prior. The low consulting a rheumatologist may be related to the low percentage response rate in the Italian study may be related to an (18%) of correct RA diagnosis made by non-rheumatologist underestimation of the RA prevalence [10], since patients with physicians in our study and possibly to a low level of public RA could have been unwilling to participate in a mail survey.
awareness of RA. Delayed rheumatological care may have On the other hand, genetic and/or environmental factors could tremendous consequences on the outcome of the disease.
account for the higher prevalence ($1.0%) in the USA [22, 23], Indeed, logistic regression showed a significant negative associa- the high prevalence of RA in Native American populations tion between early rheumatological care and ACR functional (up to 6.8%) [1], the low prevalence in Asian countries ($0.3%) classes II–IV. DMARD combination therapy was exclusively [1, 24], the rarity of RA in Africans [1], and the lack of RA in prescribed by rheumatologists and it is of interest that a significant negative association was also found between this therapy and Female sex and age !40 yrs were strong independent predictors ACR functional classes II–IV. Therefore, the early and aggressive for the disease, in our study. With the exception of a Swedish treatment prescribed by rheumatologists may account for study [5], the preponderance of RA in females is well documented the above findings. The advantages of rheumatological vs Prevalence and management of RA in Greece non-rheumatological care with regard to the outcome of the 2. MacGregor AJ, Riste LK, Hazes JMW, Silman AJ. Low prevalence disease have already been stressed [29]. Concerning the correct of rheumatoid arthritis in Black-Caribbeans compared with Whites in diagnosis and treatment of RA, the results of the non- inner city Manchester. Ann Rheum Dis 1994;53:293–7.
rheumatologist physicians were disappointing in our material, as 3. Hakala M, Po¨lla¨nen R, Nieminen P. The ARA 1987 revised criteria compared with rheumatologists; we have recently published select patients with clinical rheumatoid arthritis from a population similar findings concerning patients with seronegative spondy- based cohort of subjects with chronic rheumatic diseases registered for loarthropathies [30]. The rheumatologists had correctly diagnosed drug reimbursement. J Rheumatol 1993;20:1674–8.
and properly treated all the RA patients. About 88% of the 4. Saraux A, Guedes C, Allain J et al. Prevalence of rheumatoid arthritis patients had taken DMARDs and this is a slightly higher and spondylarthropathy in Brittany, France. J Rheumatol 1999; percentage than that reported in studies from Spain (72%) [11], France (82.1%) [31] and Canada (84%) [13]. In the present study, 5. Simonsson M, Bergman S, Jacobsson LTH, Petersson IF, Svensson B.
methotrexate was by far the most commonly employed DMARD The prevalence of rheumatoid arthritis in Sweden. Scand J Rheumatol for RA, as in other European studies [31, 32].
There may be a risk of selection bias in population-based 6. Kvien TK, Glenna˚s A, Knudsrød OG, Smedstad LM, Mowinckel P, studies. Since the participation rate in our study was high (82.1%), Førre Ø. The prevalence and severity of rheumatoid arthritis in Oslo.
selection bias is only a remote possibility. Furthermore, analysis of the data of a random sample of non-responders 7. Carmona L, Villaverde V, Herna´ndez-Garcı´a C et al. The prevalence indicated no significant difference from responders with respect to of rheumatoid arthritis in the general population of Spain.
age, sex and prevalence of rheumatic symptoms or disease.
Logistic regression showed that the random selection and non- 8. Drosos AA, Alamanos I, Voulgari PV et al. Epidemiology of adult selection of suburban and rural populations had no effect on the rheumatoid arthritis in northwest Greece 1987–1995. J Rheumatol The data on the prevalence and management of RA at the 9. Stojanovic´ R, Vlajinac H, Pablic´-Obradovic´ D, Janosˇevic´ S, level of the general adult population presented in this article were derived directly from one-to-one interviews and clinical and Yugoslavia. Br J Rheumatol 1998;37:729–32.
laboratory evaluation of the study participants by rheumatolo- 10. Cimmino MA, Parisi M, Moggiana G, Mela GS, Accardo S.
gists. The studied regions were located in northern, central Prevalence of rheumatoid arthritis in Italy: the Chiavari study. Ann and southern mainland Greece and their adult population was representative of the total Greek adult population in terms of 11. Carmona L, Gonza´lez-A´lvaro I, Balsa A et al. Rheumatoid arthritis age and sex distribution. Therefore, the results of this study could in Spain: occurrence of extra-articular manifestations and estimates of reasonably be considered as representative of the general adult disease severity. Ann Rheum Dis 2003;62:897–900.
12. Zink A, Listing J, Klindworth C, Zeidler H. The national database of the German Collaborative Arthritis Centers: I. Structure, aims, and In conclusion, our findings indicate that the prevalence of RA patients. Ann Rheum Dis 2001;60:199–206.
in the adult general population of Greece is quite similar to that in 13. Lacaille D, Anis AH, Guh DP, Esdaile JM. Gaps in care for many other European countries. Early consultation with a rheumatologist and DMARD combination therapy are associated with a better RA outcome in terms of global functional status.
14. Andrianakos A, Trontzas P, Christoyannis F et al. Prevalence of rheumatic diseases in Greece: a cross-sectional population-basedepidemiological study in urban, suburban and rural adult popula-tions. The ESORDIG study. J Rheumatol 2003;30:1589–601.
15. Arnett FC, Edworthy SM, Bloch DA et al. The Americam Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24.
17. Hochberg MC, Chang RW, Dwosh I, Lindsey S, Pincus T, Wolfe F.
The American College of Rheumatology 1991 revised criteria for the classification of global functional status in rheumatoid arthritis.
18. World Health Organisation. Obesity—Preventing and managing the global epidemic, report of a WHO consultation on obesity.
Geneva, Switzerland: World Health Organization; 1997. WHO/NUT/NCD/98.1.
19. Symmons D, Turner G, Webb R et al. The prevalence of rheumatoid We are grateful to the inhabitants, the mayors and the local arthritis in the United Kingdom: new estimates for a new century.
authorities of the studied areas for their friendly cooperation and 20. Guillemin F, Saraux A, Guggenbuhl P et al. Prevalence of rheumatoid arthritis in France—2001. Ann Rheum Dis 2005;64:1427–30.
21. Andrianakos A, Trontzas P, Voudouris C. Epidemiology of rheumatic The authors have declared no conflicts of interest.
diseases in Greece: authors reply. J Rheumatol 2004;31:1670–1.
22. Lawrence RC, Helmick CG, Arnett FC et al. Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 1998;41:778–99.
1. Silman AJ. Rheumatoid arthritis. In: Silman AJ, Hochberg MC, eds.
23. Gabriel SE, Crowson CS, O’Fallon WM. The epidemiology of Epidemiology of the Rheumatic Diseases, 2nd edn. New York: rheumatoid arthritis in Rochester, Minnesota, 1955–1985. Arthritis Oxford University Press, 2001;31–71.
24. Akar S, Birlik M, Gurler O et al. The prevalence of rheumatoid 29. Yelin EH, Such CL, Criswell LA, Epstein WV. Outcomes for persons arthritis in an urban population of Izmir-Turkey. Clin Exp with rheumatoid arthritis with a rheumatologist versus a non- rheumatologist as the main physician for this condition. Med Care 25. Minaur N, Sawyers S, Parker J, Darmawan J. Rheumatic disease in an Australian aboriginal community in north Queensland, Australia.
30. Trontzas P, Andrianakos A, Miyakis S et al. Seronegative spondy- A WHO-ILAR COPCORD survey. J Rheumatol 2004;31:965–72.
loarthropathies in Greece: a population-based study of prevalence, 26. Alarco´n GS. Epidemiology of rheumatoid arthritis. Rheum Dis Clin clinical pattern and management. The ESORDIG study. Clin 27. Gabriel SE, Crowson CS, Kremers HM et al. Survival in rheumatoid 31. Sany J, Bourgeois P, Saraux A et al. Characteristics of patients arthritis. A population-based analysis of trends over 40 years.
with rheumatoid arthritis in France: a study of 1109 patients managed by hospital based rheumatologists. Ann Rheum Dis 28. Chou C-T, Pei L, Chang D-M, Lee C-F, Schumacher HR, Liang MH.
32. Aletaha D, Smolen JS. The rheumatoid arthritis patient in the clinic: study of urban, suburban, rural differences. J Rheumatol 1994; A 40-yr-old man presented with fever, flank pain, epistaxis,haemoglobin 12.8 g/dl, WCC 12.9 Â 103/mm3, C-reactive protein(CRP) 142 mg/l and erythrocyte sedimentation rate (ESR)101 mm. Computed tomography (CT) of abdomen showed aleft renal mass, suggesting renal cell carcinoma (RCC) (Fig. 1A)and possible metastases (Fig. 1B) on CT thorax. Followingdiscussions, it was agreed that the radiological featureswere atypical for RCC, and renal abscess was more likely.
Patient remained unwell after 6 weeks of antibiotics. CRP was FIG. 1. Initial CT showing the renal mass and pulmonary nodule 320 mg/l and ESR 124 mm. No organisms grew on blood/urine culture. cytoplasmic-Anti-neutrophil cytoplasmic antibody was1:320 with strongly positive anti-PR3, suggesting Wegener’s granulomatosis (WG). Biopsy of the renal mass was planned.
Pre-biopsy CT abdomen confirmed considerable reduction in the size of the mass and new lesions in both the kidneys (Fig. 2A)compatible with vasculitis. Repeat CT thorax showed newperibronchial shadowing (Fig. 2B). Renal function deterioratedacutely and decision was made to treat for WG. Dramaticimprovement was noted. He remains well. Full blood count,renal function and CRP are normal.
Maguire et al. [1] reported atypical radiological findings in 31 WG patients; only one had a renal mass. Spontaneousresolution of the mass makes our case unique. We believe thatthe mass represented oedema surrounding the underlying FIG. 2. Repeat CT with marked reduction in the renal mass and vasculitis. Renal biopsy, while important, should not delay new renal and pulmonary lesions (marked with arrows).
treatment if the overall picture is suggestive of WG.
Correspondence to: Dr A. Negi, Specialist Registrar, The authors have declared no conflicts of interest.
Department of Rheumatology, University Hospital of Wales,Heath Park, Cardiff CF14 4XW, UK.
EGI , J. P. CAMILLERI , P. N. MATTHEWS , M. D. CRANE 1Department of Rheumatology, 2Department of Urology and3Department of Radiology, University Hospital of Wales, 1. Maguire R, Fauci AS, Doppman JL, Wolff SM. Unusual radio- graphic features of Wegener’s granulomatosis. Am J Roentgenol

Source: http://www.elire.gr/dyn/neolea/z.pdf