Radiographic sacroiliitis develops predictably over time in a cohort of familial spondyloarthritis followed longitudinally

RHEUMATOLOGY Rheumatology 2017;56:811 817 doi:10.1093/rheumatology/kew496 Advance Access publication 4 February 2017 Original article Radiographic sacroiliitis develops predictably over time in a cohort of familial spondyloarthritis followed longitudinally Félicie Costantino 1,2,3,4, Nadine Zeboulon 4, Roula Said-Nahal 4 and Maxime Breban 1,2,3,4 Abstract Objectives. Radiographic sacroiliitis is an important outcome in SpA and is considered a hallmark to ascertain the diagnosis of AS. The aim of the current study was to investigate factors associated with the presence of radiographic sacroiliitis at baseline and the predictors of progression to AS in a family cohort of SpA. Methods. A total of 953 patients fulfilling the Assessment of SpondyloArthritis international Society criteria for SpA and having at least one first- or second-degree SpA-affected relative were included. Pelvic X-rays were examined blindly and independently by two qualified examiners using the modified New York criteria. Of the 446 cases without definite sacroiliitis at inclusion, 145 patients were followed up with new pelvic X-rays for 3 15 years. Regression analysis was used to assess factors associated with definite radiographic sacroiliitis. Results. Factors independently associated with radiographic sacroiliitis at inclusion were male sex, younger age at disease onset, longer disease duration, inflammatory back pain, uveitis and lack of enthesitis. During the follow-up, 27.3% of the patients with axial SpA developed definite sacroiliitis, whereas there was no progression in patients with peripheral SpA. After 15 years of follow-up, a Kaplan Meier estimate of the proportion of patients with definite radiographic sacroiliitis reached 68.5%. Factors associated with progression to definite sacroiliitis were a low-grade radiographic sacroiliitis at inclusion, occurrence of buttock pain and the absence of peripheral arthritis during the follow-up period. Conclusions. These data confirm that progression to radiographic disease occurs most often over time in axial SpA patients. Key words: ankylosing spondylitis, spondyloarthritis, radiographic sacroiliitis, undifferentiated spondyloarthritis, non-radiographic axial spondyloarthritis Rheumatology key messages. Progression to radiographic disease occurs most often over time in axial SpA patients.. Radiographic sacroiliitis development is very rare in peripheral SpA. 1 Faculty of Health Sciences Simone Veil, 2 INSERM U1173, University of Versailles Saint-Quentin-en-Yvelines Montigny-le-Bretonneux, 3 Sorbonne Paris Cité, Laboratoire d Excellence, Université Paris Diderot, Paris and 4 Rheumatology Division, Ambroise Paré Hospital (AP-HP), Boulogne-Billancourt, France Submitted 8 July 2016; revised version accepted 19 December 2016 Correspondence to: Maxime Breban, Rheumatology Division, Hôpital Ambroise Paré, 9 ave Charles de Gaulle, 92100, Boulogne, France. E-mail: maxime.breban@aphp.fr Introduction SpA is a frequent form of chronic inflammatory arthritis with a lifetime risk recently estimated at 0.43% in the adult Caucasian population [1]. It comprises several closely related disorders characterized by axial and/or peripheral joint inflammation, often in association with extraarticular inflammation of the eye, skin or gut. Although each of these entities is defined by specific CLINICAL SCIENCE! The Author 2017. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Félicie Costantino et al. characteristics, they share several major clinical and imaging features, as well as a genetic predisposition [2]. Definite radiographic sacroiliitis is mandatory to fulfil the modified New York classification criteria for AS [3]. However, it may take up to several years from the onset of clinical symptoms to develop a definite radiographic sacroiliitis, leading to possible diagnosis delay [4]. To facilitate earlier identification of patients, the Assessment of SpondyloArthritis international Society (ASAS) developed new classification criteria for SpA in which radiographic sacroiliitis is not a mandatory criterion [5, 6]. Nevertheless, radiographic sacroiliitis remains a key component of the imaging arm of the ASAS criteria for axial SpA, along with sacroiliitis on MRI, according to which axial SpA can be divided into non-radiographic and radiographic axial SpA (i.e. AS) [5]. Several studies have compared non-radiographic and radiographic axial SpA regarding epidemiology [1, 7, 8], clinical characteristics [9, 10], disease burden [11] and response to treatment [11 15]. They support the concept of axial SpA as one entity, and non-radiographic axial SpA is often considered as an early stage of the disease, implying that sacroiliitis as ascertained on X-rays is developing over time [16]. To date, few data are available on the rate of progression from non-radiographic to radiographic axial SpA and even less on the factors associated with such progression. The aim of the current study was to investigate factors associated with the presence of radiographic sacroiliitis at baseline in a family cohort of SpA and the predictors of progression to AS during follow-up in a large fraction of the patients without definite radiographic sacroiliitis at inclusion. Methods Ethics statement This study was approved by the local ethics committees of Cochin Hospital (Paris, France) and Ambroise Paré Hospital (Boulogne-Billancourt, France). Written informed consent was obtained from each participant. Study population SpA patients included in this study belong to a family cohort originally constituted for genetic study purpose. This cohort consists of cases with at least one first- or second-degree SpA-affected relative, as previously described [2]. The diagnosis of SpA was originally ascertained according to the internationally validated classification criteria of Amor et al. and/or the European Spondyloarthropathy Study Group [17, 18]. Even if they were not developed at the time of recruitment, the ASAS classification criteria for axial and peripheral SpA were applied a posteriori [5, 6]. All the patients included in the genetic study between 1998 and 2013 with an anteroposterior radiograph of the pelvis available at inclusion were included in the crosssectional study. Grading of sacroiliitis was performed according to the established scoring system of the modified New York criteria [3] by two qualified examiners (R.S.N., M.B.). Definite radiographic sacroiliitis corresponded to a grade 52 bilateral or grade 53 unilateral. Patients who did not have definite sacroiliitis at inclusion and that could be contacted again to repeat pelvic radiographs after variable times (at least 3 years) were included in the follow-up study. For this part of the study, the same two examiners scored all the images at once (baseline and follow-up), independent of each other and without knowledge of the time order. Intra- and interobserver agreement for the two readers (k values) was previously estimated in the range of 0.86 1.0 and 0.63 0.70, respectively [19]. Discrepancies between them were resolved by consensus. Statistical analysis Cross-sectional study The association between definite radiographic sacroiliitis and clinical characteristics was examined with univariate logistic regression analysis. Baseline variables included in the analyses were the following: sex, age at disease onset, disease symptoms duration, HLA-B27 positivity, inflammatory back pain, buttock pain, peripheral arthritis, peripheral enthesitis, uveitis, psoriasis and IBD. Multivariate logistic regression models entered candidate variables (P 4 0.1 in univariate analysis). Variables retained in the final multivariate model were selected by a backward procedure. An exploratory analysis was first performed on a single randomly selected SpA case per family, in order to avoid statistical bias due to the shared genetic background within family. Moreover, to avoid bias in patient selection, this operation was repeated 10 000 times. Results of this analysis were not different from those obtained by pooling all the family SpA cases together (supplementary Table S1, available at Rheumatology Online). Thus the results are shown for the latter analysis. Longitudinal study Differences in baseline characteristics between the nonradiographic SpA patients who were followed-up in the longitudinal study and those who were not were assessed as follows: Student s t-test was used for continuous variables and chi-square test for categorical variables. Survival curves were estimated using the Kaplan Meier method. Time 0 was defined as the date of the first pelvic X-ray. The end point for analysis was the date of the first X-ray demonstrating definite radiographic sacroiliitis. Patients who did not have radiographic sacroiliitis of this degree during the follow-up were censored on the date of the last pelvic X-ray. The association with progression to definite sacroiliitis study was assessed by Cox proportional hazards analysis. Univariate Cox proportional hazard regression models were performed for a priori determined predictors. Any predictors significant at an a level of P < 0.10 were entered into a multivariate Cox proportional hazards regression analysis. All the analyses were conducted using R software (R Project for Statistical Computing, Vienna, Austria). 812 www.rheumatology.oxfordjournals.org

Factors associated with radiographic sacroiliitis in SpA TABLE 1 Clinical characteristics of the family cohort at baseline SpA without radiographic sacroiliitis Characteristics All SpA (n = 953) AS (n = 507) All (n = 446) Follow-up a (n = 145) P-value b Age at onset, mean (S.D.), years 24.6 (9.6) 23.2 (8.4) 26.1 (10.6 26.7 (9.5) 0.35 Disease duration, mean (S.D.), years 18.5 (13.0) 21.5 (13.0) 15.0 (12.0) 15.6 (12.8) 0.78 Sex, male, % 51.7 64.7 37.0 37.2 0.90 HLA-B27 positivity, % 89.1 95.1 82.2 84.8 0.48 Axial manifestations, % Inflammatory back pain 89.4 93.5 82.9 87.6 0.47 Buttock pain 86.1 90.7 80.9 80.7 0.84 Radiographic sacroiliitis c 53.2 100 0 0 - Peripheral manifestations, % Peripheral arthritis 36.0 35.7 36.2 35.9 0.35 Peripheral enthesitis 67.2 60.7 74.7 69.7 0.23 Extra-articular manifestations, % Uveitis 25.6 33.5 14.5 13.1 0.41 Psoriasis 23.6 21.4 26.1 29.0 0.55 IBD 5.7 6.1 6.6 4.1 0.26 ASAS Classification criteria fulfilment, % Axial 92.1 99.8 83.2 91.0 0.007 Peripheral 7.9 0.2 16.8 9.0 The registered manifestations correspond to those present at the time of examination or retrieved from past-medical history. Values in bold are significant values (P < 0.05). a Refers to the subset of patients who were followed up longitudinally. b P-values of the statistical comparison between patients without radiographic sacroiliitis at inclusion, whether followed up or not longitudinally. c Refers to radiographic sacroiliitis 5grade 2 bilateral or grade 3 unilateral. Results Cross-sectional study A total of 953 patients have been included in the crosssectional study. Detailed clinical characteristics of the patients are provided in Table 1. At inclusion, 53.2% of patients had definite radiographic sacroiliitis, 92.1% fulfilled ASAS classification criteria for axial SpA and 7.9% fulfilled those for peripheral SpA. We first studied axial and peripheral SpA together. Lower age at disease onset, longer disease duration, male sex, HLA-B27, inflammatory back pain, buttock pain, uveitis and lack of enthesitis were independently associated with definite radiographic sacroiliitis at inclusion (Table 2). Except for HLA-B27, factors associated with definite radiographic sacroiliitis were the same if we considered only patients fulfilling ASAS classification criteria for axial SpA (Table 3). Follow-up study Pelvic X-ray and clinical examination on follow-up were obtained for 145 of the 446 patients who did not have definite sacroiliitis at inclusion (Fig. 1). The population included in the follow-up study was well representative of all the patients without radiographic sacroiliitis. The only significant difference between the two populations was a higher frequency of axial SpA in the follow-up cohort (Table 1). The delay between inclusion and final pelvic X-ray ranged from 3 to 15 years, with a median follow-up duration of 8.3 years (interquartile range 6.7 10.9). None of the 13 patients with peripheral SpA that were followed up for 6.7 years on average exhibited radiographic progression. In contrast, 27% of the patients with non-radiographic axial SpA developed definite radiographic sacroiliitis (i.e. AS). In the latter group, Kaplan Meier estimates of the proportion of patients with definite radiographic sacroiliitis were 24.6% (95% CI 14.9, 33.2) and 68.5% (95% CI 43.6, 82.4) after 10 and 15 years of follow-up, respectively (Fig. 2). It is noteworthy that 6 of the 58 patients (10%) with lowgrade radiographic sacroiliitis at inclusion had a decreased grade at follow-up (4 patients initially graded 1a were subsequently graded 0, 1 patient graded 2a was graded 0 and 1 patient graded 2a was graded 1a). Factors associated with progression from non-radiographic axial SpA to AS in multivariate analysis were low-grade radiographic sacroiliitis at inclusion and buttock pain during the follow-up, whereas arthritis onset during the follow-up was associated with a decreased risk of developing radiographic sacroiliitis (Table 4). Discussion Since the introduction of the ASAS criteria for axial SpA, there has been considerable interest in determining whether patients identified as having non-radiographic www.rheumatology.oxfordjournals.org 813

Félicie Costantino et al. TABLE 2 Factors associated with definite radiographic sacroiliitis at inclusion in the family cohort Multivariate analysis Characteristic Univariate analysis P-value OR (95% CI) P-value Age at onset, mean (S.D.), years 4.65 10 6 0.97 (0.96, 0.99) 0.002 Disease duration, mean (S.D.), years 1.9 10 13 1.03 (1.02, 1.05) 2.0 10 7 Sex, male, % <2 10 16 3.12 (2.33, 4.35) 8.1 10 14 HLA-B27 positivity, % 1.2 10 9 2.98 (1.76, 5.05) 4.9 10 5 Articular manifestations, % Inflammatory back pain 1.2 10 5 1.80 (1.10, 2.97) 0.02 Buttock pain 5.0 10 5 2.88 (1.83, 4.53) 5.2 10 6 Peripheral arthritis 0.6 ND ND Peripheral enthesitis 2.7 10 5 0.50 (0.36, 0.68) 1.7 10 5 Extra-articular manifestations, % Uveitis 1.7 10 13 2.78 (1.94, 4.00) 3.1 10 8 Psoriasis 0.05 IBD 0.48 ND ND Values in bold are significant values (P < 0.05). ND: note done. TABLE 3 Factors associated with definite radiographic sacroiliitis at inclusion in patients with axial SpA Characteristic Radiographic sacroiliitis (n = 507) axial SpA represent those with early AS or are a group with a distinct nosology and prognosis [20]. Studying the development of radiographic sacroiliitis over time in patients with non-radiographic axial SpA can help to answer this question. In this study we estimated the rate of development of definite radiographic sacroiliitis according to the modified New York criteria in 145 patients who were followed up longitudinally for 3 15 years. This subset was highly representative of the 446 cases of familial SpA without radiographic sacroiliitis at inclusion. After a mean follow-up of nearly 9 years, we found that none of the patients fulfilling ASAS criteria for peripheral SpA progressed to AS, whereas 27.3% of those with non-radiographic axial No radiographic sacroiliitis (n = 372) Univariate analysis P-value Multivariate analysis P-value OR (95% CI) Age at onset, mean (S.D.), years 23.2 (8.4) 25.9 (10.1) 2.6 10 5 0.003 0.97 (0.96, 0.99) Disease duration, mean (S.D.), years 21.5 (13.0) 14.8 (12.7) 3.1 10 13 2.0 10 7 1.04 (1.02, 1.05) Sex, male, % 64.7 38.7 7.4 10 14 2.1 10 12 3.03 (2.22, 4.17) HLA-B27 positivity, % 95.1 95.7 0.815 ND - Axial manifestations, % Inflammatory back pain 93.5 84.9 6.6 10 5 0.05 1.68 (1.01, 2.79) Buttock pain 90.7 82.5 0.0009 1.9 10 5 2.78 (1.74, 4.45) Peripheral manifestations, % Peripheral arthritis 35.7 33.3 0.42 ND - Peripheral enthesitis 60.7 71.0 0.002 0.0004 10 4 0.56 (0.4, 0.77) Extra-articular manifestations, % Uveitis 33.5 15.9 1.9 10 10 1.0 10 7 2.71 (1.88, 3.92) Psoriasis 21.4 23.7 0.13 ND - IBD 6.1 5.1 0.20 ND - Values in bold are significant values (P < 0.05). SpA developed definite radiographic sacroiliitis. To the best of our knowledge, this study represents one of the largest in terms of the number of patients and duration of follow-up examining the rate of progression to AS in patients fulfilling the ASAS classification criteria for SpA and lacking definite radiographic sacroiliitis at inclusion. In previous similar studies the reported rates of progression ranged from 5 to 20% [21 25]: 5% in the Devenir des Spondylarthrites Indifférenciées Récentes cohort (449 patients; 2 years of follow-up), 10% in the German Spondyloarthritis Inception cohort (210 patients; 2 years of follow-up), 17.6% in a population-based cohort study from the USA (83 patients; mean follow-up of 10 years), 18.3% in the ASAS cohort (357 patients; 5 years of 814 www.rheumatology.oxfordjournals.org

Factors associated with radiographic sacroiliitis in SpA FIG. 1Flow chart of the longitudinal analysis FIG. 2Kaplan Meier curve of progression from nonradiographic axial SpA to AS follow-up) and 20% in a population-based study from Norway (20 patients; mean follow-up of 8 years). Interestingly, two of the recent studies reported regression of definite radiographic sacroiliitis on longitudinal follow-up in a proportion of cases [21, 24]. Whether similar events occurred in our cohort could not be answered, since the follow-up did not include patients having met definite radiographic sacroiliitis at inclusion. Noteworthy, however, is the fact that 10% of the patients with lowgrade radiographic sacroiliitis at inclusion had a lower grade at follow-up, suggesting a possible improvement. However, accurate ascertainment of subtle sacroiliitis changes is questionable, which hampers drawing definite conclusions about the reality of such radiographic regression in our setting. Older studies performed before the introduction of the ASAS classification criteria used a variety of inclusion criteria. Sampaio-Barros et al. [26] showed that 24.3% of the patients with undifferentiated SpA progressed to AS over 5 10 years. Schattenkirchner and Krüger [27] reported a 25% progression rate after a follow-up period ranging from 2 to 6 years in patients with B27-positive oligoarthritis. Progression rates to AS were higher in other reports by Mau et al. [28] (59% after 10 years), Huerta-Sil et al. [29] (42% after a mean follow-up of 3 years) and Bennett et al. [30] (33% after a mean follow-up of 7 years), but those estimates were based on relatively few SpA patients (ranging from 33 to 50 per study). Discrepancies between several studies could presumably be explained by differences in inclusion criteria, disease duration at baseline, duration of follow-up and genetic and/or environmental factors that might affect disease severity. The x-axis represents the years of follow-up; the y-axis represents the proportion remaining as non-radiographic axial SpA. Numbers below the plot represent the number of subjects at risk at each time. Sixteen years ago we described the baseline characteristics of a subset of the family cohort examined herein, representing roughly one-third of those patients [2]. In that report, we observed that the frequency of radiographic sacroiliitis increased from 40% in patients with disease durations of <10 years to 86% in patients with disease durations of 520 years. We had thus inferred that the non-radiographic forms of SpA likely represented early forms of AS, although such an assumption required a prospective follow-up to be confirmed. This was the main objective of the present study, results of which are fully consistent with our earlier interpretation, by showing that as many as two-thirds of the patients without radiographic sacroiliitis at baseline have evolved towards full-blown AS over 15 years of evolution. We also identified several factors robustly associated with the presence of definite radiographic sacroiliitis in both the entire cohort and the subset of patients fulfilling ASAS axial SpA criteria at inclusion: lower age at disease onset, longer disease duration, male sex, inflammatory back pain, uveitis and lack of enthesitis. An influence of disease duration on radiographic sacroiliitis has already been reported [2, 31]. Other studies also reported a significantly higher male:female ratio in radiographic as compared with non-radiographic axial SpA [10, 12, 22]. In contrast, previously published studies failed to identify an association of radiographic sacroiliitis with any specific SpA clinical feature, possibly because of a lower statistical power than in the present study [10, 12, 22]. A high level of CRP has also been associated with radiographic axial SpA by others [10, 12, 22, 31]. Here we did not collect www.rheumatology.oxfordjournals.org 815

Félicie Costantino et al. TABLE 4 Factors associated with progression to definite radiographic sacroiliitis in non-radiographic axial SpA cases Univariate Multivariate Characteristic a HR (95% CI) P-value HR (95% CI) P-value Age at onset 0.98 (0.05, 1.02) 0.36 ND ND Disease duration 1.00 (0.97, 1.03) 0.81 ND ND Sex 1.17 (0.59, 2.32) 0.65 ND ND Present at the time of examination or retrieved from past medical history Axial manifestations Inflammatory back pain 1.09 (0.45, 2.64) 0.85 ND ND Buttock pain 2.53 (0.97, 6.63) 0.06 2.06 (0.69, 6.19) 0.20 Low grade radiographic sacroiliitis 1.38 (1.15, 1.67) 0.0007 1.46 (1.19, 0.78) 0.0003 Peripheral manifestations Peripheral arthritis 0.55 (0.27, 1.11) 0.10 1.01 (0.42, 2.39) 0.98 Peripheral enthesitis 0.94 (0.46, 1.96) 0.88 ND ND Extra-articular manifestations Uveitis 1.73 (0.75, 3.98) 0.20 ND ND Psoriasis 0.87 (0.41, 1.87) 0.73 ND ND IBD 0.96 (0.13, 7.06) 0.97 ND ND During follow-up Axial manifestations Inflammatory back pain 1.12 (0.53, 2.39) 0.77 ND ND Buttock pain 3.22 (0.99, 10.53) 0.05 3.38 (1.01, 11.33) 0.05 Peripheral manifestations Peripheral arthritis 0.36 (0.17, 0.78) 0.009 0.34 (0.14, 0.80) 0.01 Peripheral enthesitis 0.78 (0.51, 1.55) 0.51 ND ND Extra-articular manifestations Uveitis 1.49 (0.69, 3.22) 0.32 ND ND Psoriasis 0.99 (0.48, 2.06) 0.98 ND ND IBD 0.25 (0.03, 1.85) 0.17 ND ND Values in bold are significant values (P < 0.05). ND: not done. a HLA-B27 could not be included as a variable since all the nonradiographic SpA cases under study were HLA-B27 positive. such information systematically and were thus unable to analyse the influence of CRP level in our study. We found that low-grade radiographic sacroiliitis at inclusion and buttock pain or lack of arthritis during the followup were independent predictors of radiographic sacroiliitis progression. Low-grade radiographic sacroiliitis and buttock pain, but not arthritis, have already been associated with the progression to AS in previous studies [26, 29]. Among other factors previously associated with the progression to AS, we did not observe the influence of HLA- B27, a finding that can be readily explained by the relatively few HLA-B27-negative patients in the familial context. In conclusion, a progression to AS was observed in 27.3% of the 132 patients with non-radiographic axial SpA at inclusion followed-up over a mean period of nearly 9 years. Moreover, the Kaplan Meier curve allowed us to estimate that this proportion would possibly reach 68.5% after 15 years of follow-up. These data confirm that progression to radiographic disease occurs most often over time in axial SpA patients. It reinforces the assumption that at least in the family context, all forms of SpA represent a single multifaceted disorder rather than a panel of distinct, albeit related, conditions [20]. The clinical significance of progression of radiographic sacroiliitis remains unanswered and it would be interesting to collect more longitudinal data to know if such progression is associated with physical function. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article. Disclosure statement: The authors have declared no conflicts of interest. Supplementary data Supplementary data are available at Rheumatology Online. References 1 Costantino F, Talpin A, Said-Nahal R et al. Prevalence of spondyloarthritis in reference to HLA-B27 in the French population: results of the GAZEL cohort. Ann Rheum Dis 2015;74:689 93. 2 Said-Nahal R, Miceli-Richard C, Berthelot JM et al. The familial form of spondylarthropathy: a clinical study of 115 multiplex families. Groupe Français d Etude Génétique des Spondylarthropathies. Arthritis Rheum 2000;43: 1356 65. 816 www.rheumatology.oxfordjournals.org

Factors associated with radiographic sacroiliitis in SpA 3 van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984;27:361 8. 4 Rudwaleit M, van der Heijde D, Khan MA et al. How to diagnose axial spondyloarthritis early. Ann Rheum Dis 2004;63:535 43. 5 Rudwaleit M, van der Heijde D, Landewé R et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68:777 83. 6 Rudwaleit M, van der Heijde D, Landewé R et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general. Ann Rheum Dis 2011;70:25 31. 7 Reveille JD, Witter JP, Weisman MH. Prevalence of axial spondylarthritis in the United States: estimates from a cross-sectional survey. Arthritis Care Res 2012;64:905 10. 8 Strand V, Rao SA, Shillington AC et al. Prevalence of axial spondyloarthritis in United States rheumatology practices: Assessment of SpondyloArthritis International Society criteria versus rheumatology expert clinical diagnosis. Arthritis Care Res 2013;65:1299 306. 9 Rudwaleit M, Haibel H, Baraliakos X et al. The early disease stage in axial spondylarthritis: results from the German Spondyloarthritis Inception Cohort. Arthritis Rheum 2009;60:717 27. 10 Kiltz U, Baraliakos X, Karakostas P et al. Do patients with non-radiographic axial spondylarthritis differ from patients with ankylosing spondylitis? Arthritis Care Res 2012;64: 1415 22. 11 Boonen A, Sieper J, van der Heijde D et al. The burden of non-radiographic axial spondyloarthritis. Semin Arthritis Rheum 2015;44:556 62. 12 Ciurea A, Scherer A, Exer P et al. Tumor necrosis factor a inhibition in radiographic and non-radiographic axial spondyloarthritis: results from a large observational cohort. Arthritis Rheum 2013;65: 3096 106. 13 Song IH, Weiss A, Hermann KG et al. Similar response rates in patients with ankylosing spondylitis and nonradiographic axial spondyloarthritis after 1 year of treatment with etanercept: results from the ESTHER trial. Ann Rheum Dis 2013;72:823 5. 14 Sieper J, Landewe R, Rudwaleit M et al. Effect of certolizumab pegol over ninety-six weeks in patients with axial spondyloarthritis: results from a phase III randomized trial. Arthritis Rheumatol 2015;67:668 77. 15 Wallman JK, Kapetanovic MC, Petersson IF et al. Comparison of non-radiographic axial spondyloarthritis and ankylosing spondylitis patients baseline characteristics, treatment adherence, and development of clinical variables during three years of anti-tnf therapy in clinical practice. Arthritis Res Ther 2015;17:378. 16 Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis: do we need new criteria? Arthritis Rheum 2005;52: 1000 8. 17 Amor B, Dougados M, Mijiyawa M. [Criteria of the classification of spondylarthropathies]. Rev Rhum Mal Osteoartic 1990;57:85 9. 18 Dougados M, van der Linden S, Juhlin R et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991;34:1218 27. 19 Said-Nahal R, Miceli-Richard C, D Agostino MA et al. Phenotypic diversity is not determined by independent genetic factors in familial spondylarthropathy. Arthritis Rheum 2001;45:478 84. 20 Baeten D, Breban M, Lories R et al. Are spondylarthritides related but distinct conditions or a single disease with a heterogeneous phenotype? Arthritis Rheum 2013;65: 12 20. 21 Dougados M, Demattei C, van den Berg R et al. Rate and predisposing factors of sacroiliac radiographic progression after a 2 years follow-up period in recent onset spondyloarthritis. Arthritis Rheumatol 2016;68:1904 13. 22 Poddubnyy D, Rudwaleit M, Haibel H et al. Rates and predictors of radiographic sacroiliitis progression over 2 years in patients with axial spondyloarthritis. Ann Rheum Dis 2011;70:1369 74. 23 Wang R, Gabriel SE, Ward MM. Progression of patients with non-radiographic axial spondyloarthritis to ankylosing spondylitis: a population-based cohort study. Arthritis Rheumatol 2016;68:1415 21. 24 Sepriano A, Rudwaleit M, Sieper J et al. Five-year followup of radiographic sacroiliitis: progression as well as improvement? Ann Rheum Dis 2016;75:1262 3. 25 Bakland G, Alsing R, Singh K et al. Assessment of SpondyloArthritis International Society criteria for axial spondyloarthritis in chronic back pain patients with a high prevalence of HLA-B27. Arthritis Care Res 2013;65: 448 53. 26 Sampaio-Barros PD, Bortoluzzo AB, Conde RA et al. Undifferentiated spondyloarthritis: a longterm followup. J Rheumatol 2010;37:1195 9. 27 Schattenkirchner M, Krüger K. Natural course and prognosis of HLA-B27-positive oligoarthritis. Clin Rheumatol 1987;6(Suppl 2):83 6. 28 Mau W, Zeidler H, Mau R et al. Outcome of possible ankylosing spondylitis in a 10 years follow-up study. Clin Rheumatol 1987;6(Suppl 2):60 6. 29 Huerta-Sil G, Casasola-Vargas JC, LondoñoJDet al. Low grade radiographic sacroiliitis as prognostic factor in patients with undifferentiated spondyloarthritis fulfilling diagnostic criteria for ankylosing spondylitis throughout follow up. Ann Rheum Dis 2006;65:642 6. 30 Bennett AN, McGonagle D, O Connor P et al. Severity of baseline magnetic resonance imaging-evident sacroiliitis and HLA-B27 status in early inflammatory back pain predict radiographically evident ankylosing spondylitis at eight years. Arthritis Rheum 2008;58:3413 8. 31 Poddubnyy D, Brandt H, Vahldiek J et al. The frequency of non-radiographic axial spondyloarthritis in relation to symptom duration in patients referred because of chronic back pain: results from the Berlin early spondyloarthritis clinic. Ann Rheum Dis 2012;71:1998 2001. www.rheumatology.oxfordjournals.org 817