Abstract
Objectives:
To investigate effect of therapies on radiographic progression in axial spondyloarthritis (axSpA).
Methods:
Comprehensive database search for studies assessing radiographic progression in axSpA (particular treatment vs. no treatment of interest) was performed. Study-specific standardized mean differences were estimated and combined using random-effects model.
Results:
Twenty four studies were included; 18 with tumor necrosis factor inhibitors (TNFi), 8 with non-steroidal anti-inflammatory drugs (NSAIDs), and 1 with secukinumab. Spinal radiographic progression was not significantly different among TNFi-treated vs. biologic naïve populations at 2 years (mSASSS difference= −0.73, 95% CI −1.52 to 0.12, I2=28%) and at ≥4 years (mSASSS difference= −2.03, 95% CI −4.63 to 0.72, I2=63%). Sensitivity analysis restricted to studies with low risk of bias showed a significant difference at ≥4 years (mSASSS difference= −2.17, 95% CI −4.19 to −0.15). No significant difference was observed between NSAIDs vs. control (mSASSS difference= −0.30, 95% CI =−2.62 to 1.31, I2=71%), or secukinumab vs. biologic naïve (mSASSS difference= −0.34, 95% CI −0.85 to 0.17). There were not enough studies on nr-axSpA or SIJ progression for analysis.
Conclusions:
Although no significant protective effect of TNFi treatment on spinal radiographic progression of AS at 2 and ≥4 years was seen, analysis restricted to studies with low risk of bias showed a protective effect at ≥4 years. Therefore, long-term TNFi exposure might have radiographic progression benefit. No difference was seen with NSAIDs or secukinumab at 2 years. Future studies should explore effect of biologics on radiographic progression in early axSpA and nr-axSpA, and with long-term exposure.
Keywords: spondyloarthritis, ankylosing spondylitis, radiographic progression, TNF inhibitor, NSAIDs
INTRODUCTION
Treatment goals for axial spondyloarthritis (axSpA), including ankylosing spondylitis (AS) and non-radiographic axSpA (nr-axSpA), are control of spinal and peripheral symptoms as well as prevention of radiographic damage [1]. Radiographic progression takes place in 20–45% of AS patients after 2 years and progression is expected to continue in successive years, leading to complete fusion of spine (“bamboo spine”) in up to 40% of patients with AS [2,3]. Rate of progression is slower in nr-axSpA, about 7% at 2 years [3].
Conventional radiography is the gold standard for assessment of radiographic progression [4], although magnetic resonance imaging can detect active inflammation before changes due to chronic damage appear on radiograph [5]. Assessment of SpondyloArthritis international Society(ASAS) and Outcome Measures in Rheumatology (OMERACT) Working Groups recommend use of modified Stoke AS Spine Score (mSASSS) for quantitative assessment of axial damage [6,7]. mSASSS ranges from 0 to 72, signifying structural changes on lateral radiographs of cervical and lumbar spine [8]. It that has the highest reliability and greatest validation compared to Bath AS Radiology Index (BASRI) and sacroiliac joint (SIJ) scores [6,7,9]. The minimum interval for significant radiographic change is 2 years [10,11], and a change of 2 mSASSS units in 2 years or 1 new syndesmophyte formation in 2 years is considered as radiographic progression [2,12].
Although radiographic progression is an important predictor of poor functional outcomes in axSpA [13,14], treatment is largely symptomatic at present. If any therapy is shown to slow the natural progression of disease, it might support introducing them early with treat-to-target strategy, similar to rheumatoid arthritis (RA). The current guidelines recommend against this due to lack of evidence [1].
Considering the clinical implications of radiographic progression and conflicting study results, we aimed to perform a systematic review and meta-analysis on the effect of different therapies on radiographic progression in axSpA.
METHODS
Search strategy and study selection
A comprehensive search of several databases from inception to January 15, 2019 was conducted. The databases included Ovid Medline In-Process & Other Non-Indexed Citations, Ovid MEDLINE, Ovid EMBASE, Ovid Cochrane un Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus. ACR and EULAR abstracts indexed in MEDLINE were included as well. Search strategy was designed and conducted by a medical reference librarian with input from the principal investigator (Supplementary File 1).
We included all original reports fulfilling the following criteria:
Adults (≥18 years) with axSpA, including AS and nr-axSpA.
Participants were assessed for radiographic progression of axial disease (outcome) using a quantitative scoring method (mean mSASSS score, number of syndesmophytes, SIJ score, or any other scoring method).
Radiographic scores were reported with respect to a particular therapy and compared to placebo or a group without the therapy of interest.
Duration of follow-up at least 1 year.
While cohort and case-control studies were included; cross-sectional studies, case series, case reports, and non-human studies were excluded. The effect of combination therapy was beyond the scope of this study.
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting systematic reviews [15] was followed (Figure 1). Two authors (PK and AD) screened abstracts for eligibility, retrieved full-texts and excluded irrelevant articles. The inter-rater agreement was very good (kappa statistic- 0.91, 95% CI- 0.81 to 0.99). Disagreements were resolved by discussion about eligibility. Bibliographies belonging to included studies, reviews, and relevant articles were screened for additional studies. No language restrictions were made. Language translation was done by translators proficient both in the particular language and in English (AD and EM).
Figure 1.
Flow chart describing systematic search and study selection process
Data extraction/handling methods
Relevant data were extracted by PK, and checked by RS. Data from available full texts were considered in preference to conference abstracts, unless the abstracts contained updated results. Where a matched data analysis was provided (for example, in studies with historical comparison group), this was extracted with preference to data on all participants. Data on radiographic progression were collected at 2 and ≥4 years of follow up. Authors were contacted for additional data where necessary.
Outcomes
The main outcome was difference in mSASSS in AS between groups at 2 years and ≥4 years. Other radiographic scores such as: mSASSS in nr-axSpA, change in number of syndesmophytes, BASRI-spine, CT score of the facet joints and SIJ radiographic progression in AS and/or nr-axSpA were secondary outcomes. We did not pool mSASSS with SIJ or other scores in our analysis as pooling different scoring systems has not been validated. For example, the SIJ score is semi-quantitative with relatively poor correlation with functional status [16] and inter-reader reliability [17].
The effect size measured as the standard mean difference (SMD) of the change in radiographic score from baseline between treatment and control group was used.
SMD was estimated as:
SMD= (Radiographic score T, baseline− Radiographic score T, follow-up) − (Radiographic score C, baseline− Radiographic score C, follow-up) divided by SD Pooled
Where SD Pooled is the pooled SD of the radiographic change in treatment (T) and control groups (C) at baseline and follow-up.
Statistical analysis
All outcome comparisons and treatment effects were calculated using RevMan version 5.3. Binary measures were converted to SMD (SE) [18]. Study specific SMDs were estimated and combined using the random-effects model as described by DerSimonian and Laird[19]. Rescaling to original units (eg. mSASSS and number of syndesmophytes) was done by multiplying SMD by SD to allow better clinical interpretation. The SD was taken from the average of the pooled SD of change in radiographic scores (of the treatment and control group) from several trials reporting the original scale [20].
Radiographic outcomes in AS and nr-axSpA were reported separately and not pooled. To maintain independence, only 2 of 4 studies with the Outcome Assessment in Ankylosing Spondylitis (OASIS) cohort as control (with the largest number of patients) were included in 2 separate forest plots (one study each for 2 years [21] and >4 years [22]). Sensitivity analysis was carried out with other studies [23,24], one at a time. Similarly, 2 studies from the Prospective Study Of AS (PSOAS) cohort were included only in separate analyses [25,26]. Data from Gensler et al. [26] was used in preference to an older publication from PSOAS cohort (mSASSS was defined as mSASSS of ≥1 unit/year, and variable follow-up periods of 1.5 to 9 years) [27]. Sample mean and SD were calculated from sample size, median, range and/or interquartile range [28].
Assessment of risk of bias, certainty in the evidence and heterogeneity
Studies were independently evaluated by PK and RS for risk of bias [29]. New-Castle Ottawa scale [30] was used for cohort studies, and Revised Cochrane risk-of-bias tool (RoB 2.0) was used for randomized trials [31] (Supplementary File 2). The certainty in the evidence was evaluated using the GRADE approach (Grading of Recommendations, Assessment, Development and Evaluation) methodology [32]. Publication bias was assessed visually using funnel plots (Supplementary File 3). Between-study heterogeneity was assessed using I2 statistics (i.e., I2 < 30% - low heterogeneity, 30 to 60%- moderate, and >60% - high) [33].
RESULTS
Characteristics of the included studies
Out of 524 studies screened, 24 studies (23 English, 1 Russian language) fulfilled our inclusion criteria: 18 studies related to TNFi (N= 4,874), 8 to NSAIDs (N= 2,321), and 1 to secukinumab (N=237). Among these, 3 studies contained data for both NSAIDs and TNFi [25,26,34]. Included studies were mostly cohort studies and open label extensions of randomized controlled trials (RCTs). Three of the TNFi studies were excluded from the meta-analysis (details provided above in the statistical analysis section). One RCT for TNFi [35], and 2 RCTs for NSAIDs [36,37] were noted (Table 1). Most studies were judged to have low risk of bias, except 2 observational studies [38,39] and 2 RCTs [36,37] (Supplementary File 2). One study on NSAIDs [40] was excluded as no quantitative data was available and phenylbutazone (no longer approved for human use in US) was studied.
Table 1.
Baseline characteristics of the included studies
| Author, Year | Study Design | Follow-up duration (years) | AS/ AxSpA characteristics | Treatment | Control | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Drug | N | Disease duration, yrs (SD) | Baseline mSASSS (SD) | Drug | N | Disease duration, yrs (SD) | Baseline mSASSS (SD) | ||||
| TNFi | |||||||||||
| Baraliakos 2005 [42] | Open label extension of RCT | 2 | Active AS | Infliximab 5 mg/kg IV every 6 wk | 41 | 15.5 (7.5) | 12.1 (16.9) | No TNFi, GESPIC historical cohort | 41 | 5.5 (2.25) | 5.9 (13.4) |
| Baraliakos 2007 [22] | Open label extension of RCT | 4 | Active AS (same cohort as 2005 study) | Infliximab 5 mg/kg IV every 6 wk | 33 | 19.0 (23.4) | 11.6 (15.3) | No TNFi, OASIS historical cohort | 132 | 11.7 (9.3) | 12.7 (17.4) |
| van der Heijde (I) 2008 [24] | Open label extension of 24 week RCT | 2 (96 wk) |
Active AS (ASSERT cohort) | Infliximab 5 mg/kg IV every 6 wk after loading dose | 201 | 10.2 (8.7) | 17.7 (17.9) | No TNFi, OASIS historical cohort-matched | 70 | 9.9 (8.8) | 17.5 (19.1) |
| van der Heijde (E) 2008 [23] | Open label extension of 24 week RCT | 2 (96 wk) |
AS | Etanercept 25 mg SC twice a wk | 257 | 10 (8.5) | 14 (17.6) | No TNFi, OASIS historical cohort (meeting RCT entry criteria) | 76 | 12 (9.8) | 19 (20.8) |
| van der Heijde 2009 [21] | Open label extension of 24 week RCT | 2 | AS (Canadian [M03–606] study and the ATLAS study group | Adalimumab 40 mg SC every other week | 307 | 19.8 (19.3) | 19.8 (19.3) | No TNFi, OASIS historical cohort (eligible patients) | 77 | 11.3 (8.7) | 15.8 (17.6) |
| Pedersen 2011 [45] | Cohort study | 2 | AS | Infliximab 3 or 5 mg/kg (11), etanercept 25 mg twice a wk (10), adalimumab 40 mg every other wk (2) | 23 | 18.2 (11.4) | 14.5 916.1) | No TNFi, standard therapy | 27 | 15 (10) | 10.0 (12.1) |
| Haroon 2013 [27] | Cohort, prospective | 1.5 to 9 | AS | TNFi- type and dose not specified | 201 | - | - | No TNFi | 133 | - | - |
| Kang 2013 [46] | Cohort study | 2 | AS | Infliximab, etanercept or adalimumab | 26 | 9.5 (5.1) | 4.0 (6.6) | No TNFi ( NSAID and/or MTX or SSZ) | 37 | 8.0 (4.5) | 3.7 (6.8) |
| Min 2014 [38] | Cohort, retrospective, single-center | 8 | AS | TNFi- type and dose not specified | 14 | - | - | Continuous NSAIDs | 12 | - | - |
| Braun 2014 [35] | Phase 3, multicentric, randomized, placebo-controlled, double- blind, placebo crossover | 2 (104 wk) |
Active AS (GO-RAISE trial) | Golimumab 50 or 100 mg every 4 wk | 233 | 7.25 (35.59) | 12.64 (17.71) | Placebo, crossover to golimumab 50 mg at wk 16 or 24 | 66 | 5.20 (45.94) | 16.1 (18.7) |
| 4 | Active AS (GO-RAISE trial) | Golimumab 50 or 100 mg every 4 wk | 233 | 7.25 (35.59) | 12.64 (17.71) | Placebo, crossover to golimumab 50 mg at wk 16 or 24 | 66 | 5.20 (45.94) | 16.1 (18.7) | ||
| Baraliakos 2014 [43] | Open label extension of RCT | 8 | AS (DIKAS) | Infliximab 5 mg/kg IV every 6 wk | 22 | 15.8 (8.5) | 13.2 (17.6)- adjusted to 13.8 for comparison | No TNFi, Herne historical cohort | 34 | 20.7 (5.7) | 14.2 (13.8)- adjusted to 13.8 for comparison |
| Chung 2015 [47] | Clinical trial | 4 (42– 66 mo) | AS | TNFi- type and dose not specified | 25 | 4.15 (4.02) | - | No TNFi | 25 | 2.13 (1.73) | - |
| Minhas 2016 [25] | Cohort, prospective | ≥2, up to >10 (Median= 3) | AS (PSOAS cohort) | TNFi (>50% of FU)- type and dose not specified | 630 (total) | - | - | No TNFi (<50% of FU period) | 630 (total) | - | - |
| Kim 2016 [48] | Cohort, prospective | 5 | AS (OKSAR cohort) | TNFi- type and dose not specified | 269 | 11.33 (7.51) | 18.87 (17.96) | TNF naïve | 341 | 8.04 (6.57) | 15.68 (15.49) |
| Dougados 2018 [41] | Open label extension of RCT | 2 | Nr-axSpA (EMBARK trial) | Etanercept 25 mg SC twice a wk | 162 | 2.4 (1.8) | Baseline total SIJ score- 1.5 (1.2) | No biologics, Contemporary DESIR cohort | 193 | 1.70 (1.0) | Baseline SIJ score= 1.9 (1.6) |
| Molnar 2018 [34] | Cohort study | 10 (2 year radiographic interval progression) | AS (patients fulfilling modified NY criteria for AS from SCQM AxSpA cohort) | Any TNFi before radiographic interval, NSAIDs | 163 | 13.8 (9.7) | 6.6 (12.5) | No TNFi before radiographic interval | 269 | - | - |
| Gensler 2018 [26] | Cohort, prospective | 2, 4 | AS (PSOAS cohort) | TNFi- type and dose not specified | 239 | 16.8 (12.5) | 14.2 (19.6) | No TNFi | 280 | 16.8 (12.5) | 14.2 (19.6) |
| Park 2019 [44] | Cohort study, single center | 4 | Early AS (<10 year symptom duration) | TNFi- type and dose not specified | 135 | 2.7 (2.6) | 6.2 (9.9) | NSAIDs, Control group from different institution (same location-Seoul) | 80 | 0.7 (1.8) | 7.3 (10.8) |
| NSAIDs | |||||||||||
| Wanders 2005 [37] | RCT, open label, radiographs blinded | 2 | AS patients | Continuous NSAIDs (started on celecoxib 200 mg bid but allowed to increase or change NSAID) | 76 | 13 (10.2) | 7.9 (14.7) | On-demand NSAIDs ( celecoxib or another NSAID) | 74 | 10.2 (9.3) | 9.3 (15.2) |
| Poddubnyy 2012 [49] | Cohort study | 2 | AxSpA (GESPIC) | High NSAID intake (NSAID intake ≥50) | 43 | - | - | Low NSAID intake (NSAID index <50) | 121 | - | - |
| AS | 24 | 5.5 (2.7) | 6.7 (7.7) | 64 | 5.0 (2.9) | 5.7 (11.6) | |||||
| Nr-AxSpA | 19 | 3.7 (2.1) | 1.6 (4.0) | 57 | 3.0 (2.2) | 2.6 (4.8) | |||||
| Gensler 2018 [26] | Cohort, prospective | 2,4 | AS (PSOAS cohort) | NSAIDs | 343 | 16.8 (12.5) | 14.2 (19.6) | No NSAIDs | 176 | 16.8 (12.5) | 14.2 (19.6) |
| Schiotis 2013 [50] | Cohort study | 3 | AS (REGISPONSER), BASRI spine ≥12 excluded | Continuous NSAIDs | 81 | 25.32 (9.38) | BASRI spine- 7.23 (3.34) | On-demand NSAIDs | 37 | 25.54 (12.3) | BASRI-spine= 6.69 (3.4) |
| Minhas 2016 [25] | Cohort, prospective | >2, up to >10 (Median= 3) | AS (PSOAS cohort) | NSAID index>50 | total in both groups=630 | - | - | - | total in both groups=630 | - | - |
| Sieper 2016 [36] | RCT, open label, radiographs blinded | 2 | AS (ENRADAS) | Continuous Diclofenac 150 mg /day (or equivalent dose if could not tolerate) | 85 | 12.2 (10.3) | 11.3 (14.9) | On-demand NSAIDs | 82 | 15.2 (12.4) | 14.0 (16.8) |
| Molnar 2018 [34] | Cohort study | 10 (2 year radiographic interval progression) | AS (patients fulfilling modified NY criteria for AS from SCQM AxSpA cohort) | NSAIDs | 286 | 13.8 (9.7) | 6.6 (12.5) | no NSAIDs | 55 | - | - |
| Rumyantseva 2018 [39] | Cohort study | 2 | Early AxSpA | Continuous NSAIDs | 35 | 24.1 (15.4) | SIJ score- 3.67 (0.74) | On-demand NSAIDs | 33 | 24.1 (15.4) | 3 (1.48) |
| Secukinumab | |||||||||||
| Braun 2018 [51] | Retrospective analysis of RCT | 2 | AS (MEASURE 1) | Secukinumab 150 mg or 75 mg IV every 4 wk | 168 | - | 9.55 (14.14) | Biologic-naïve (on NSAIDs only), ENRADAS | 69 | - | 9.95 (13.76) |
AS- ankylosing spondylitis (defined as meeting modified New York criteria), axSpA- axial spondyloarthritis, nr- non-radiographic, N- number of participants, wk-week, SD- standard deviation, mSASSS- modified Stoke AS Spine Score, TNFi- tumor necrosis factor alpha, NSAIDs- non-steroidal anti-inflammatory drugs, BASRI(Bath AS Radiology Index), SIJ- sacroiliac joint, OASIS -Outcome Assessment in Ankylosing Spondylitis, ASSERT-Ankylosing Spondylitis Study for the Evaluation of Recombinant Infliximab Therapy, ATLAS- Adalimumab Trial Evaluating Long‐term Efficacy and Safety for Ankylosing Spondylitis, SCQM- Swiss Clinical Quality Management, GESPIC- early German AS cohort, DIKAS- Seutsche Infliximab Kohorte fur AS, GO-RAISE- golimumab in ankylosing spondylitis, PSOAS- Prospective Study of Outcomes in Ankylosing Spondylitis, GLAS- Groningen Leeuwarden AS, ENRADAS- Effects of NSAIDs on Radiographic Damage in Ankylosing Spondylitis, DESIR- French Cohort of Undifferentiated Spondyloarthritis, REGISPONSER- Spanish National Registry of Spondyloarthropathies.
Among TNFi studies, there were 17 studies in AS and 1 study in nr-axSpA [41]. Six studies used a historical cohort as a comparator, not on TNFi (OASIS, 4 studies [20,31–33]; the German Spondyloarthritis Inception cohort (GESPIC), 1 study [42]; Herne cohort, 1 study [43]). Two studies used contemporary cohorts, not on TNFi as comparators [41,44]. The control group was on continuous NSAIDs [38], NSAIDs [44], or standard of care with no TNFi. The only RCT for TNFi used placebo as control until 24 weeks, after which they crossed over to receive golimumab 50 mg (same as the intervention arm) through 100 weeks [35]. The type of TNFi used was variable: 3 studies used infliximab [22,24,42], 2 studies etanercept [23,41], 1 study adalimumab [21], 1 study golimumab [35], 2 studies used etanercept, infliximab and adalimumab [45,46], and type of TNFi was not specified in remainder of the 7 studies. Only one of the studies compared 2 doses of TNFi (golimumab), and none of the other studies compared TNFi dose response in relation to radiographic progression [35]. Maximum duration of follow-up was 2 years in 7 studies [20,32–34,38–40], ≥ 4 years in 9 studies [22,28,31,35–37,42–44], and one study had a median follow-up of 3 years ( >2 years to >10 years) [25]. The mean duration of disease was >5 years in most studies, except in 3 of the included studies [41,44,47]. Baseline mSASSS varied between studies and also within study groups, ranging from 4 to 18.87 (median= 13.20) in the TNFi group and 3.70 to 19 (median=14.20) in the control group.
Among 8 NSAID studies, there were 6 studies with only AS patients, 1 study with early axSpA [39], and 1 study with both AS and nr-axSpA [49]. Six studies reported mSASSS [26,34,36,37,39,49], 1 study BASRI spine [50], and 1 study BASRI-SIJ [25]. The maximum duration of follow-up was 2 years in 5 studies and >2 years in 3 of the studies [25,34,50]. Baseline mSASSS ranged from 6.60 to 14.20 (median= 7.90) in the NSAID group, and 5.70 to 14.20 (median=11.65) in the control group in AS patients. In one of the studies reporting nr-axSpA, the mean mSASSS was 1.60 (SD=4) compared to 2.60 (SD=4.80) [49].
The only included study (abstract) with secukinumab was a retrospective analysis of an RCT, comparing 2 year data to the historical biologic-naïve Effects of NSAIDs on Radiographic Damage in Ankylosing Spondylitis (ENRADAS) cohort [51]. mSASSS was comparable between the 2 groups: 9.55 (SD=14.14) in the secukinumab group and 9.95 (SD=13.76) in the control group. The full manuscript was published beyond the inclusion period for our review, and data from the abstract were verified with the final published version (June 2019) [52]. The original trial and 4 year data were not included in our study because while two different doses of secukinumab were compared, the study did not have a NSAID or placebo arm [53,54].
Radiographic outcomes- TNFi
In assessment of spinal radiographic progression with TNFi in AS patients, most studies reported mSASSS (15 studies) whereas one study reported a CT score of facet joints [47] (Table 2). No studies on spinal radiographic progression with TNFi in nr-axSpA were found. Five studies also measured change in number of syndesmophytes [34,38,43,45,46]. Spinal radiographic progression was not significantly different among the TNFi-treated vs. the biologic-naïve populations at 2 years (mSASSS difference= −0.73, 95% CI −1.52 to 0.12, I2=28%) and at ≥4 years (mSASSS difference= −2.03, 95% CI −4.63 to 0.72, I2=63%). (Figure 2A & B). However, sensitivity analysis restricted to six studies with low risk of bias (excluding one study[38]) showed a significant difference at ≥4 years (mSASSS difference= −2.17, 95% CI −4.19 to −0.15, I2=49%) (Figure 2C). Certainty in this following the GRADE approach was low (supplemental file 2).
Table 2.
Radiographic progression at the spine and sacroiliac joint in the included studies
| Author, Year | Follow-up duration (years) | Treatment | Control | Difference between groups | ||||
|---|---|---|---|---|---|---|---|---|
| N | Change in mSASSS (SD) | No of new syndesmophytes (SD), other radiographic parameters | N | Change in mSASSS (SD) | No of new syndesmophytes (SD), other radiographic parameters | |||
| TNFi | ||||||||
| Baraliakos 2005 [42] | 2 | 41 | 0.4 (2.7) | - | 41 | 0.7 (2.8) | - | - |
| Baraliakos 2007 [22] | 4 | 33 | 1.6 (2.6) | - | 132 | 4.4 (26.21) | - | - |
| van der Heijde (I) 2008 [24] | 2 | 201 | 0.9 (2.6) | - | 70 | 1.2 (3.9) | - | - |
| van der Heijde (E) 2008 [23] | 2 | 257 | 0.91 (2.45) | Change in Cervical radiography score- 0.49 (1.40), Lumbar radiography score- 0.42 (1.84) | 76 | 1.27 (3.64) | Change in Cervical radiography score- 0.53 (2.29), Lumbar radiography score- 0.73 (2.00) | - |
| van der Heijde 2009 [21] | 2 | 307 | 0.8 (2.6) | - | 77 | 0.9 (4.1) | - | - |
| Pedersen 2011 [45] | 2 | 23 | 1.4 (1.9) | 0.52 (0.8) | 27 | 1.5 (3.1) | 0.70 (1.4) | - |
| Haroon 2013 [27] | >1.5 to 9 | 201 | - | - | 133 | - | - | mSASSS OR- 0.52 (0.30 to 0.88) |
| Kang 2013 [46] | 2 | 26 | 3.3 (4.2) | 0.9 (1.4) | 37 | 2.3 (5.1) | 0.4 (0.9) | - |
| Min 2014 [38] | 8 | 14 | 9.29 (6.22) | 2.79 (2.94) | 12 | 4.58 (2.15) | 0.50 (0.674) | - |
| Braun 2014 [35] | 2 | 233 | 0.9 (3.33) | - | 66 | 1.6 (4.6) | - | - |
| 4 | 233 | 1.67 (4.89) | - | 66 | 2.1 (5.2) | - | - | |
| Baraliakos 2014 [43] | 2 | 22 | 3.2 (7.04) | - | 34 | 2.7 (9.33) | - | - |
| 4 | 4.4 (6.57) | - | - | 17.5 (10.49) | - | - | ||
| 6 | 6.2 (6.57) | - | - | 19.3 (7.58) | - | - | ||
| 8 | 7.2 (6.57) | 1 (8.64) | - | 11.70 (6.41) | 2.70 (6.25) | - | ||
| Chung 2015 [47] | 4 | 25 | - | Change in CT score of facet joints, total spine- 11.72 (53.94) | 25 | - | Change in CT score of facet joints, total spine- 11.56 (43.02) | - |
| Minhas 2016 [25] | >2, up to >10 | 630 | - | - | - | - | - | BASRI (radiographic change in SI joint) OR- 0.06 (0.004 to 0.99) |
| Kim 2016 [48] | 5 | 269 | 4.73 (18.56) | - | 341 | 6.14 (32.80) | - | - |
| Dougados 2018 [41] | 2 | 162 | - | Change in total SIJ score= −0.14 (0.81) | 193 | - | Change in total SIJ score= 0.08 (0.83) | -0.22 (95% CI −0.38,−0.06) |
| Molnar 2018 [34] | 10 (2 year radiographic interval progression) | 163 | - | - | 269 | - | - | mSASSS OR- 0.50 ( 95% CI 0.28 to 0.88), progression of ≥1 new syndesmophytes per 2 years- OR 0.55 (0.33 to 0.94) |
| Gensler 2018 [26] | 2 | 239 | - | - | 280 | - | - | Mean mSASSS difference= −0.71 (−2.10 to 0.68, p=0.32) |
| 4 | 239 | - | - | 280 | - | - | Mean mSASSS difference= −1.37 (−2.07 to −0.63 p<0.001) | |
| Park 2019 [44] | 4 | 135 | - | - | 80 | - | - | OR - 0.41 (95% CI 0.22 to 0.75) |
| NSAIDs | ||||||||
| Wanders 2005 [37] | 2 | 76 | 0.4 (1.7) | - | 74 | 1.5 (2.5) | - | - |
| Poddubnyy 2012 [49] | 2 | 43 | - | - | - | |||
| AS | 24 | 0.02 (1.39) | - | 64 | 0.96 (2.78) | - | - | |
| Nr-axSpA | 19 | 0.51 (1.72) | - | 57 | 0.74 (1.95) | - | - | |
| Gensler 2018 [26] | 2 | 343 | - | - | 176 | - | - | Mean mSASSS difference= 0.64 (−1.52 to 2.81, p=0.56) |
| 4 | 343 | - | - | 176 | - | - | Mean mSASSS difference= 1.95 (0.90 to 2.99, p <0.001) | |
| Schiotis 2013 [50] | 3 | 81 | - | Change in BASRI spine- 0.64 (1.22) | 37 | - | Change in BASRI spine- 0.66 (1.04) | - |
| Minhas 2016 [25] | >2, up to >10 | 630 | - | - | - | - | BASRI SIJ OR- 0.03 (0.002 to 3.50) | |
| Sieper 2016 [36] | 2 | 85 | 1.29 (2.82) | - | 82 | 0.71 (2.77) | - | - |
| Molnar 2018 [34] | 10 (2 year radiographic interval progression) | 286 | - | - | 55 | - | - | OR- 0.85 (0.42 to 1.72), progression of ≥1 new syndesmophytes per 2 years- OR 1.04 (0.50 to 2.17) |
| Rumyantseva 2018 [39] | 2 | 35 | - | Change in SIJ score= 1 (1.66) | 33 | - | Change in SIJ score-=1.33 (2.67) | - |
| Secukinumab | ||||||||
| Braun 2018 [51] | 2 | 168 | 0.55 (1.82) | % with no radiographic progression 61% | 69 | 0.89 (1.83) | % with no radiographic progression 52% | - |
AS- ankylosing spondylitis, Nr-axSpA- non-radiographic axial spondyloarthritis, N- number of participants, SD- standard deviation, mSASSS- modified Stoke AS Spine Score, TNFi- tumor necrosis factor alpha, NSAIDs- non-steroidal anti-inflammatory drugs, BASRI- Bath AS Radiology Index-spine, SIJ- sacroiliac joint, CT- computed tomography.
Figure 2.
Forest plot of radiographic progression in patients on tumor necrosis factor inhibitor alpha (TNFi) A) radiographic progression at the spine in AS- 2 years, B) radiographic progression at the spine in AS- ≥ 4 years, C) sensitivity analysis with low risk of bias studies for radiographic progression at the spine in AS- ≥ 4 years.
Conversion to mSASSS was done by multiplying the SMD by the average pooled SD (6.06 at 2 years [21,26,35,42,45,46], and 14.47 at 4 years [22,26,35,38,43,48]) as explained in the methods section. Only one of the studies reported change in CT score of facet joints, which was not significantly different between TNFi and biologic-naïve groups (CT score difference= −0.16, 95% CI −27.91 to 27.59) [47]. The largest study with OASIS as the control group [21] was included in primary analysis (one each in 2 years and ≥4 years follow-up). Sensitivity analyses performed with other studies that used OASIS as control, one at a time, did not show any difference. Subgroup analyses performed to assess heterogeneity did not reveal significant differences between historical and contemporary control groups (p=0.54). Similarly, there was no difference in the number of syndesmophytes between the TNFi and biologic-naïve groups at 2 years (SMD= −0.04, 95% CI −0.51 to 0.43; change in no. of syndesmophytes= −0.05, 95% CI −0.59 to 0.49, I2=69%) or 8 years of follow-up (SMD= 0.34, 95% CI −0.86 to 1.55; change in number of syndesmophytes= 0.78, 95% CI −3.01 to 4.57, I2=83%). Average pooled SD (1.15 at 2 years [45,46] and 2.29 at 8 years [38,43]) from studies reporting a change in the number of syndesmophytes was used for the conversion [46].
Only 2 studies [25,41] reported radiographic changes at the SIJ. Minhas et al. reported OR= 0.06 (95% CI= 0.004 to 0.99) for BASRI- SIJ (radiographic change in SI joint) in the PSOAS cohort for the TNFi vs. biologic-naïve groups [25]. Dougados et al. reported change in total SIJ score of −0.22 (95% CI −0.38 to −0.06, p=0.008) in TNFi vs. biologic-naïve groups in nr-axSpA [41]. Pooled estimates were not calculated for the 2 studies as the study populations were different (AS vs. nr-axSpA) (Figure 3B).
Figure 3.
Forest plot of radiographic progression in patients on tumor necrosis factor inhibitor alpha (TNFi) A) change in the number of syndesmophytes, and B) radiographic progression at sacroiliac joint (SIJ) in AS and nr-axSpA respectively.
Radiographic outcomes- NSAIDs
Among 6 studies reporting mSASSS with NSAIDs in AS at 2 years, no significant difference was observed between NSAIDs vs. control group (SMD= −0.08, 95% CI −0.32 to 0.16, mSASSS difference= −0.30, 95% CI =−2.62 to 1.31, I2=71%)[26,34,36,37,49] (Figure 4). Average pooled SD (8.18) was used for the conversion [26,36,37]. Dosing strategies for NSAIDs in both treatment and control arms were different- comparisons were made between continuous vs. on-demand [36,37], NSAID index >50 vs. <50 [49] and NSAID vs. no NSAID [26,34]. Subgroup analysis (continuous vs on-demand NSAIDs use, NSAID index- high vs low, and NSAIDs vs no NSAIDs) of these showed no difference (p=0.79). There was one study which compared BASRI-spine for NSAID vs. control, which showed no difference as well (BASRI-spine difference= 0.020, 95% CI= −0.44 to 0.48) [50]. One study reported a subgroup of patients with nr-axSpA, in whom no difference in mSASSS was observed (mSASSS difference= 0.13, 95% CI=−0.39 to 0.65) [49]. Sensitivity analysis after removing an observational study judged to have high risk of bias [39] did not change the results.
Figure 4.
Forest plot of radiographic progression in patients on non-steroidal anti-inflammatory drugs (NSAIDs) A) radiographic progression at the spine in AS at 2 years (mSASSS only), B) radiographic progression at the spine in AS at 3 years (BASRI-spine), and C) radiographic progression at SIJ in AS and early axSpA respectively.
Two studies [25,39] with NSAIDs reported changes in SIJ based on BASRI-SIJ, where no difference in SIJ radiographic progression was seen in the two groups (SMD=−0.18, 95% CI −0.65 to 0.29, SIJ score difference= −0.40, 95% CI −1.44 to 0.64, I2=0%). Pooled SD (2.21) was used for the conversion [39].
Radiographic outcomes- secukinumab
The only included study with secukinumab, did not show a significant difference in radiographic progression over 2 years (mean mSASSS difference= −0.34, 95% CI −0.85 to 0.17) [51].
DISCUSSION
This systematic review and meta-analysis showed that TNFi may slow radiographic progression at the spine in AS at ≥4 years (when only studies judged to have low risk of bias were included) but not at 2 years. Recent studies have suggested that progression might be less beyond 2 years [55]; and a non-linear, continued benefit beyond 4 years has been observed with TNFi [26,56]. These studies adjust for important confounders (eg. smoking status, NSAID use, baseline mSASSS and disease/symptom duration), which might account for the difference in results. Our study showed a larger difference in mSASSS between TNFi and biologic-naïve group at ≥4 years, suggesting continued benefit with long-term use of TNFi. The difference in radiographic progression at SIJ in AS and nr-axSpA was not significant in each of the 2 included studies (Figure 3). The semi-quantitative nature of the SIJ score with inter-reader variability, make the study of radiographic progression at SIJ challenging.
In contrast, NSAIDs did not show any significant inhibition of radiographic progression in AS at the spine at 2 years or SIJ at >2 years. A clinically significant effect at long term follow up is possible and will require long-term studies. Furthermore, differential effects of NSAIDs have been noted, and are believed to be related to the degree of Cox-2 selectivity. Wanders et al. showed that AS patients on continuous celecoxib have lower radiographic progression in comparison to on-demand treatment over 2 years, although the cumulative dose between the groups was not much different [37]. Also the study was not blinded, and may have resulted in differential use of co-interventions such as exercise, smoking cessation, etc. In a post-hoc analysis of this trial, those with elevated acute phase reactants seemed to have the greatest benefit with continuous celecoxib compared to on-demand NSAIDs [57]. It is postulated that elevated prostaglandin E2 leads to increased osteoblastic activity, and hence inhibition of prostaglandin (especially prostaglandin E2) synthesis by Cox-2 inhibitors might inhibit new bone formation [58,59]. However, a subsequent non-blinded RCT (ENRADAS trial) that used diclofenac, failed to confirm these findings [36]. It is unclear whether the differential effect on bone formation with selective Cox-2 inhibitors used in Wanders et al. as opposed to the non-selective NSAIDs used in ENRADAS trial led to the difference in results. At present, there is insufficient evidence to confirm the effect of NSAIDs (selective or non-selective) on radiographic progression alone or in combination with TNFi. Long-term risk of different NSAIDs in this population should be studied further in terms of cardiovascular and gastrointestinal safety to justify the risk-benefit profile.
The only study with secukinumab did not show a significant difference in radiographic progression at 2 years [51]. More studies are required to better understand the long-term effects and effects in early disease. Phase III data from secukinumab showed no increase in spinal radiographic damage in 80% of AS patients at 2 and 4 years [53,54], but these data need to be interpreted with caution in the absence of a control group.
A paucity of data on the effect of treatment on radiographic progression in nr-axSpA was noted. A subgroup of nr-axSpA from the GESPIC cohort showed no difference in mSASSS progression between high and low NSAID use at 2 years [49]. Only one study exploring the effects of TNFi on SIJ progression from the EMBARK trial was included in our study, which did not show any significant difference [41]. Study by Almirall et al. and the RAPID-axSpA study were not included in our study as all patients in these studies were treated with TNFi with no comparator arm [55,60]. These studies showed no radiographic progression at the spine/SIJ at 2 years and SIJ at 4 years, respectively. However, the data are difficult to interpret in the absence of a controlled comparison. No data on secukinumab or biologics with other modes of action was found.
Although our study showed a significant effect of TNFi on long-term radiographic progression (in sensitivity analysis), none of the included studies provide prospective, long-term controlled comparison. Most included studies were judged to have a low risk of bias; however predominance of observational and open-label extensions of RCTs limits overall level of evidence. Most analyses in our study showed low to moderate heterogeneity and a few (on NSAIDs) showed high heterogeneity. There were methodological differences in between the studies and various subgroup and sensitivity analyses, e.g., historical vs. contemporary controls, and dosing of NSAIDs were done. These explain some heterogeneity in the data, but not all.
Answering the question of radiographic progression will require a concerted effort. Long term, controlled trials of axSpA therapies are costly, therefore alternative strategies will be necessary to learn which therapies are best in preventing radiographic progression. Well-designed RCTs with head-to-head comparisons will be required to establish the comparative efficacy of biologic therapies, either alone or in combination with NSAIDs, in slowing radiographic progression. A head-to-head study of secukinumab with TNFi is planned [61], which will hopefully give us more information regarding their comparative efficacy. There is an ongoing RCT comparing radiographic progression at 2 years between treatment with TNFi alone and the combination of TNFi and NSAIDs, following promising data on the combination approach from observational studies [26]. There is also increasing evidence to suggest that TNFi in early AS (<10 years) is associated with a higher benefit. Observational studies by Haroon et al. [27] and Park et al. [44] both suggested the importance of early initiation of TNFi therapy on radiographic progression. Data from observational studies will be limited as the expense of serial imaging tests; and using observational data to assess radiographic progression will likely require funding specifically to obtain serial imaging at standardized intervals among patients who are treated with TNFi or IL-17 inhibitors alone or in combination with NSAIDs. Secondly, risk stratification to identify those at high risk of progression is important given that not all patients progress. Risk factors for radiographic progression of axSpA noted in multiple studies are male sex, HLA-B27 positivity, baseline radiographic changes, long disease/symptom duration, high C-reactive protein, high disease activity, smoking status, [62] and more recently alcohol has been implicated [63]. Close monitoring and early therapy with a treat-to-target strategy might have a greater impact on slowing structural damage in this high risk group. Lastly, we need more sensitive and reliable measures to document radiographic progression. While most studies used mSASSS, the most validated measure for radiographic progression in AS, it is based on plain radiographs with limited sensitivity to change. mSASSS doesn’t include assessment of changes at thoracic spine or posterior elements (facet joints), and cannot assess early damage [64,65]. Inter-reader reliability for mSASSS change is also of concern, with most studies showing poor to moderate reliability [21,24,52]. While plain radiography is cheap, fast, and has years of experience with reading; it might not be the best measure for assessment due to the slow nature of radiographic progression in AS. Newer measures, such as those based on quantitative low-dose computed tomography (ld-CT) scan may provide higher sensitivity in detecting small changes in syndesmophyte size/volume [66,67]. Ld-CT has also been shown to have good correlation with various measures of patient function such as Schober test and lateral thoracolumbar flexion [68]. Comparison of x-ray measurement (mSASSS) with these newer modalities will guide whether CT is an acceptable outcome for measure in clinical trials.
CONCLUSION
Although no significant protective effect of TNFi treatment on radiographic progression of AS at the spine at 2 years and 4 years was found in our study, analysis restricted to studies with low risk of bias showed a protective effect at ≥4 years. Therefore, long-term TNFi exposure might have radiographic progression benefit. No difference was seen with NSAIDs or secukinumab (only 1 study) at 2 years but long-term data were not available. Further studies should explore the effect of NSAIDs and biologics alone and in combination in patients with early axSpA; their use in the group with high risk of progression should be evaluated with a follow up >4 years to see if effects are more pronounced over time. Newer measures with higher sensitivity to detect structural changes, such as those based on quantitative low-dose CT should be compared to mSASSS for use in clinical trials.
Supplementary Material
Supplementary File 3. Funnel plot of included studies looking at A) radiographic progression at the spine- TNFi (2 years), B) radiographic progression at the spine- TNFi (≥ 4 years), C) change in the number of syndesmophytes- TNFi, and D) radiographic progression at the spine- NSAIDs.
Supplemental File 1. Search strategy for the systematic review process
Supplementary File 2A) Risk of bias of the included studies i) New-Castle Ottawa scale for observational cohort studies, and ii) Revised Cochrane risk-of-bias tool (RoB 2.0) for randomized controlled trials, 2B) GRADE quality assessment of main outcomes for TNFi and NSAIDs.
ACKNOWLEDGEMENT
We would like to thank our librarian, Larry Prokop for help with the extensive search strategy; Drs. Victor M. Montori and Colin P. West, Mayo Clinic for their guidance with the systematic review and meta-analysis process; Dr. Lianne S. Gensler, University of California San Francisco for her expertise and guidance with the protocol, and providing supplemental data from her abstract on PSOAS cohort.
This project was supported by CTSA Grant Number UL1 TR002377 from the National Center for Advancing Translational Science (NCATS) (PK), NIH AR0691427 (MD), AR04775 (MD), and T32 GM08685 (PS). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
Ali Duarte-Garcia is supported by the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, which receives no industry funding.
Elena Myasoedova is supported by the Gerstner Family Career Development Award from the Center of Individualized Medicine.
REFERENCES
- 1.Ward MM, Deodhar A, Gensler LS, et al. 2019 Update of the American College of Rheumatology/Spondylitis Association of America/Spondyloarthritis Research and Treatment Network Recommendations for the Treatment of Ankylosing Spondylitis and Nonradiographic Axial Spondyloarthritis. Arthritis Rheumatol Hoboken NJ Published Online First: 22 August 2019. doi: 10.1002/art.41042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Baraliakos X, Listing J, von der Recke A, et al. The natural course of radiographic progression in ankylosing spondylitis--evidence for major individual variations in a large proportion of patients. J Rheumatol 2009;36:997–1002. doi: 10.3899/jrheum.080871 [DOI] [PubMed] [Google Scholar]
- 3.Poddubnyy D, Haibel H, Listing J, et al. Baseline radiographic damage, elevated acute-phase reactant levels, and cigarette smoking status predict spinal radiographic progression in early axial spondylarthritis. Arthritis Rheum 2012;64:1388–98. doi: 10.1002/art.33465 [DOI] [PubMed] [Google Scholar]
- 4.Braun J, Golder W, Bollow M, et al. Imaging and scoring in ankylosing spondylitis. Clin Exp Rheumatol 2002;20:S-178–S-184. [PubMed] [Google Scholar]
- 5.Oostveen J, Prevo R, den Boer J, et al. Early detection of sacroiliitis on magnetic resonance imaging and subsequent development of sacroiliitis on plain radiography. A prospective, longitudinal study. J Rheumatol 1999;26:1953–8. [PubMed] [Google Scholar]
- 6.Wanders AJB, Landewé RBM, Spoorenberg A, et al. What is the most appropriate radiologic scoring method for ankylosing spondylitis? A comparison of the available methods based on the Outcome Measures in Rheumatology Clinical Trials filter. Arthritis Rheum 2004;50:2622–32. doi: 10.1002/art.20446 [DOI] [PubMed] [Google Scholar]
- 7.van der Heijde D, Landewé R. Selection of a method for scoring radiographs for ankylosing spondylitis clinical trials, by the Assessment in Ankylosing Spondylitis Working Group and OMERACT. J Rheumatol 2005;32:2048–9. [PubMed] [Google Scholar]
- 8.Creemers MCW, Franssen MJAM, van’t Hof MA, et al. Assessment of outcome in ankylosing spondylitis: an extended radiographic scoring system. Ann Rheum Dis 2005;64:127–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ramiro S, Claudepierre P, Sepriano A, et al. Which scoring method depicts spinal radiographic damage in early axial spondyloarthritis best? Five-year results from the DESIR cohort. Rheumatol Oxf Engl 2018;57:1991–2000. doi: 10.1093/rheumatology/key185 [DOI] [PubMed] [Google Scholar]
- 10.MacKay K, Mack C, Brophy S, et al. The Bath Ankylosing Spondylitis Radiology Index (BASRI): a new, validated approach to disease assessment. Arthritis Rheum 1998;41:2263–70. doi: [DOI] [PubMed] [Google Scholar]
- 11.Spoorenberg A, de Vlam K, van der Linden S, et al. Radiological scoring methods in ankylosing spondylitis. Reliability and change over 1 and 2 years. J Rheumatol 2004;31:125–32. [PubMed] [Google Scholar]
- 12.Ramiro S, Stolwijk C, van Tubergen A, et al. Evolution of radiographic damage in ankylosing spondylitis: a 12 year prospective follow-up of the OASIS study. Ann Rheum Dis 2015;74:52–9. doi: 10.1136/annrheumdis-2013-204055 [DOI] [PubMed] [Google Scholar]
- 13.Landewe R, Dougados M, Mielants H, et al. Physical function in ankylosing spondylitis is independently determined by both disease activity and radiographic damage of the spine. Ann Rheum Dis 2009;68:863–7. doi: 10.1136/ard.2008.091793 [DOI] [PubMed] [Google Scholar]
- 14.Poddubnyy D, Listing J, Haibel H, et al. Functional relevance of radiographic spinal progression in axial spondyloarthritis: results from the GErman SPondyloarthritis Inception Cohort. Rheumatol Oxf Engl 2018;57:703–11. doi: 10.1093/rheumatology/kex475 [DOI] [PubMed] [Google Scholar]
- 15.Beller EM, Glasziou PP, Altman DG, et al. PRISMA for Abstracts: Reporting Systematic Reviews in Journal and Conference Abstracts. PLoS Med 2013;10. doi: 10.1371/journal.pmed.1001419 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Exarchou S, Redlund-Johnell I, Karlsson M, et al. The prevalence of moderate to severe radiographic sacroiliitis and the correlation with health status in elderly Swedish men--the MrOS study. BMC Musculoskelet Disord 2013;14:352. doi: 10.1186/1471-2474-14-352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.van den Berg R, Lenczner G, Feydy A, et al. Agreement between clinical practice and trained central reading in reading of sacroiliac joints on plain pelvic radiographs. Results from the DESIR cohort. Arthritis Rheumatol Hoboken NJ 2014;66:2403–11. doi: 10.1002/art.38738 [DOI] [PubMed] [Google Scholar]
- 18.9.4.6 Combining dichotomous and continuous outcomes. https://handbook-5-1.cochrane.org/chapter_9/9_4_6_combining_dichotomous_and_continuous_outcomes.htm (accessed 20 Mar 2019).
- 19.DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88. doi: 10.1016/0197-2456(86)90046-2 [DOI] [PubMed] [Google Scholar]
- 20.Murad MH, Wang Z, Chu H, et al. When continuous outcomes are measured using different scales: guide for meta-analysis and interpretation. BMJ 2019;364:k4817. doi: 10.1136/bmj.k4817 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.van der Heijde D, Salonen D, Weissman BN, et al. Assessment of radiographic progression in the spines of patients with ankylosing spondylitis treated with adalimumab for up to 2 years. Arthritis Res Ther 2009;11:R127. doi: 10.1186/ar2794 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Baraliakos X, Listing J, Brandt J, et al. Radiographic progression in patients with ankylosing spondylitis after 4 yrs of treatment with the anti-TNF-alpha antibody infliximab. Rheumatology 2007;46:1450–3. doi: 10.1093/rheumatology/kem166 [DOI] [PubMed] [Google Scholar]
- 23.van der Heijde D, Landewé R, Einstein S, et al. Radiographic progression of ankylosing spondylitis after up to two years of treatment with etanercept. Arthritis Rheum 2008;58:1324–31. doi: 10.1002/art.23471 [DOI] [PubMed] [Google Scholar]
- 24.van der Heijde D, Landewe R, Baraliakos X, et al. Radiographic findings following two years of infliximab therapy in patients with ankylosing spondylitis. Arthritis Rheum 2008;58:3063–70. doi: 10.1002/art.23901 [DOI] [PubMed] [Google Scholar]
- 25.Minhas D, Lee MJ, Rahbar MH, et al. Do TNF inhibitors change the progression of sacroiliitis? Arthritis Rheumatol 2016;68:2089–90. doi: 10.1002/art.39977 [DOI] [Google Scholar]
- 26.Gensler LS, Gianfrancesco M, Weisman MH, et al. Combined effects of tumour necrosis factor inhibitors and nsaids on radiographic progression in ankylosing spondylitis. Ann Rheum Dis 2018;77 (Supplement 2):148. doi: 10.1136/annrheumdis-2018-eular.4027 [DOI] [Google Scholar]
- 27.Haroon N, Inman RD, Learch TJ, et al. The impact of tumor necrosis factor α inhibitors on radiographic progression in ankylosing spondylitis. Arthritis Rheum 2013;65:2645–54. doi: 10.1002/art.38070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wan X, Wang W, Liu J, et al. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol 2014;14:135. doi: 10.1186/1471-2288-14-135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cochrane Handbook for Systematic Reviews of Interventions. /handbook (accessed 10 Mar 2019).
- 30.Ottawa Hospital Research Institute. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp (accessed 10 Mar 2019).
- 31. https://www.riskofbias.info/welcome/rob-2-0-tool.
- 32.Murad MH. Clinical Practice Guidelines: A Primer on Development and Dissemination. Mayo Clin Proc 2017;92:423–33. doi: 10.1016/j.mayocp.2017.01.001 [DOI] [PubMed] [Google Scholar]
- 33.Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. doi: 10.1002/sim.1186 [DOI] [PubMed] [Google Scholar]
- 34.Molnar C, Scherer A, Baraliakos X, et al. TNF blockers inhibit spinal radiographic progression in ankylosing spondylitis by reducing disease activity: results from the Swiss Clinical Quality Management cohort. Ann Rheum Dis 2018;77:63–9. doi: 10.1136/annrheumdis-2017-211544 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Braun J, Baraliakos X, Hermann K-GA, et al. The effect of two golimumab doses on radiographic progression in ankylosing spondylitis: results through 4 years of the GO-RAISE trial. Ann Rheum Dis 2014;73:1107–13. doi: 10.1136/annrheumdis-2012-203075 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Sieper J, Listing J, Poddubnyy D, et al. Effect of continuous versus on-demand treatment of ankylosing spondylitis with diclofenac over 2 years on radiographic progression of the spine: Results from a randomised multicentre trial (ENRADAS). Ann Rheum Dis 2016;75:1438–43. doi: 10.1136/annrheumdis-2015-207897 [DOI] [PubMed] [Google Scholar]
- 37.Wanders A, Heijde D van der, Landewé R, et al. Nonsteroidal antiinflammatory drugs reduce radiographic progression in patients with ankylosing spondylitis: a randomized clinical trial. Arthritis Rheum 2005;52:1756–65. doi: 10.1002/art.21054 [DOI] [PubMed] [Google Scholar]
- 38.Min HK, Kang JY, Koh JH, et al. Comparisons of radiographic progression of ankylosing spondylitis between treatment with TNF antagonist, continuous treatment with NSAID, and on demand treatment of NSAID. Clin Exp Rheumatol 2014;32 (5):829. [Google Scholar]
- 39.Rumyantseva DG, Dubinina TV, Erdes SF. Impact of the frequency of using nonsteroidal anti-inflammatory drugs on the radiographic progression of sacroiliitis in patients with early axial spondyloarthritis. Nauchno-Prakt Revmatol 2018;56:346–50. doi: 10.14412/1995-4484-2018-346-350 [DOI] [Google Scholar]
- 40.Boersma JW. Retardation of ossification of the lumbar vertebral column in ankylosing spondylitis by means of phenylbutazone. Scand J Rheumatol 1976;5:60–4. [PubMed] [Google Scholar]
- 41.Dougados M, Maksymowych WP, Landewé RBM, et al. Evaluation of the change in structural radiographic sacroiliac joint damage after 2 years of etanercept therapy (EMBARK trial) in comparison to a contemporary control cohort (DESIR cohort) in recent onset axial spondyloarthritis. Ann Rheum Dis 2018;77:221–7. doi: 10.1136/annrheumdis-2017-212008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Baraliakos X, Listing J, Rudwaleit M, et al. Radiographic progression in patients with ankylosing spondylitis after 2 years of treatment with the tumour necrosis factor alpha antibody infliximab. Ann Rheum Dis 2005;64:1462–6. doi: 10.1136/ard.2004.033472 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Baraliakos X, Haibel H, Listing J, et al. Continuous long-term anti-TNF therapy does not lead to an increase in the rate of new bone formation over 8 years in patients with ankylosing spondylitis. Ann Rheum Dis 2014;73:710–5. doi: 10.1136/annrheumdis-2012-202698 [DOI] [PubMed] [Google Scholar]
- 44.Park JW, Kim MJ, Lee JS, et al. Impact of Tumor Necrosis Factor Inhibitor Versus Nonsteroidal Antiinflammatory Drug Treatment on Radiographic Progression in Early Ankylosing Spondylitis: Its Relationship to Inflammation Control During Treatment. Arthritis Rheumatol 2019;71:82–90. doi: 10.1002/art.40661 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Pedersen SJ, Chiowchanwisawakit P, Lambert RGW, et al. Resolution of inflammation following treatment of ankylosing spondylitis is associated with new bone formation. J Rheumatol 2011;38:1349–54. doi: 10.3899/jrheum.100925 [DOI] [PubMed] [Google Scholar]
- 46.Kang KY, Ju JH, Park S-H, et al. The paradoxical effects of TNF inhibitors on bone mineral density and radiographic progression in patients with ankylosing spondylitis. Rheumatology 2013;52:718–26. doi: 10.1093/rheumatology/kes364 [DOI] [PubMed] [Google Scholar]
- 47.Chung SW, Song R, Lee YA, et al. Evaluation of longitudinal association between changes in facet joints and administration of tumor necrosis factor inhibitors in ankylosing spondylitis using CT. Ann Rheum Dis 2015;74:1162. doi: 10.1136/annrheumdis-2015-eular.1357 [DOI] [Google Scholar]
- 48.Kim T-J, Shin J-H, Kim S, et al. Radiographic progression in patients with ankylosing spondylitis according to tumor necrosis factor blocker exposure: Observation Study of Korean Spondyloarthropathy Registry (OSKAR) data. Jt Bone Spine Rev Rhum 2016;83:569–72. doi: 10.1016/j.jbspin.2015.09.006 [DOI] [PubMed] [Google Scholar]
- 49.Poddubnyy D, Rudwaleit M, Haibel H, et al. Effect of non-steroidal anti-inflammatory drugs on radiographic spinal progression in patients with axial spondyloarthritis: results from the German Spondyloarthritis Inception Cohort. Ann Rheum Dis 2012;71:1616–22. doi: 10.1136/annrheumdis-2011-201252 [DOI] [PubMed] [Google Scholar]
- 50.Schiotis R, Font P, Escudero A, et al. Long term influence of NSAIDs on radiographic progression in patients with ankylosing spondylitis. Ann Rheum Dis Conf Annu Eur Congr Rheumatol Eur Leag Rheum EULAR 2013;72. doi: 10.1136/annrheumdis-2013-eular.1544 [DOI] [Google Scholar]
- 51.Braun J, Haibel H, De Hooge M, Landewé RB, Rudwaleit M, Readie A, et al. Low rate of spinal radiographic progression over 2 years in ankylosing spondylitis patients treated with secukinumab: a historical cohort comparison [abstract]. Arthritis Rheumatol 2018;70 Suppl 10 URL: https://acrabstracts.org/abstract/lowrate-of-spinal-radiographic-progression-over-2-years-in-ankylosing-spondylitis-patients-treated-with-secukinumab-a-historicalcohort-comparison/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Braun J, Haibel H, de Hooge M, et al. Spinal radiographic progression over 2 years in ankylosing spondylitis patients treated with secukinumab: a historical cohort comparison. Arthritis Res Ther 2019;21:142. doi: 10.1186/s13075-019-1911-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Braun J, Baraliakos X, Deodhar A, et al. Effect of secukinumab on clinical and radiographic outcomes in ankylosing spondylitis: 2-year results from the randomised phase III MEASURE 1 study. Ann Rheum Dis 2017;76:1070–7. doi: 10.1136/annrheumdis-2016-209730 [DOI] [PubMed] [Google Scholar]
- 54.Braun J, Baraliakos X, Deodhar A, et al. Secukinumab shows sustained efficacy and low structural progression in ankylosing spondylitis: 4-year results from the MEASURE 1 study. Rheumatology 2018;19:19. doi: 10.1093/rheumatology/key375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.van der Heijde D, Baraliakos X, Hermann K-GA, et al. Limited radiographic progression and sustained reductions in MRI inflammation in patients with axial spondyloarthritis: 4-year imaging outcomes from the RAPID-axSpA phase III randomised trial. Ann Rheum Dis 5;77:699–705. doi: 10.1136/annrheumdis-2017-212377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Maas F, Arends S, Brouwer E, et al. Reduction in Spinal Radiographic Progression in Ankylosing Spondylitis Patients Receiving Prolonged Treatment With Tumor Necrosis Factor Inhibitors. Arthritis Care Res 2017;69:1011–9. doi: 10.1002/acr.23097 [DOI] [PubMed] [Google Scholar]
- 57.Kroon F, Landewé R, Dougados M, et al. Continuous NSAID use reverts the effects of inflammation on radiographic progression in patients with ankylosing spondylitis. Ann Rheum Dis 2012;71:1623–9. doi: 10.1136/annrheumdis-2012-201370 [DOI] [PubMed] [Google Scholar]
- 58.Blackwell KA, Raisz LG, Pilbeam CC. Prostaglandins in bone: bad cop, good cop? Trends Endocrinol Metab TEM 2010;21:294–301. doi: 10.1016/j.tem.2009.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Vuolteenaho K, Moilanen T, Moilanen E. Non-steroidal anti-inflammatory drugs, cyclooxygenase-2 and the bone healing process. Basic Clin Pharmacol Toxicol 2008;102:10–4. doi: 10.1111/j.1742-7843.2007.00149.x [DOI] [PubMed] [Google Scholar]
- 60.Almirall M, Lopez-Velandia JG, Maymo J. Absence of radiographic progression at two years in a cohort of patients with non-radiographic axial spondyloarthritis treated with TNF-alpha blockers. Reumatol Clin 2014;10:134–5. doi: 10.1016/j.reuma.2013.04.003 [DOI] [PubMed] [Google Scholar]
- 61.Baraliakos X, Ostergaard M, Gensler LS, et al. Comparison of secukinumab and adalimumab biosimilar on radiographic progression in ankylosing spondylitis patients: Design of a randomized, phase-IIIb study (SURPASS). Int J Rheum Dis 2018;21 (Supplement 1):121–2. doi: 10.1111/1756-185X.13361 [DOI] [Google Scholar]
- 62.Sari I, Haroon N. Radiographic Progression in Ankylosing Spondylitis: From Prognostication to Disease Modification. Curr Rheumatol Rep 2018;20 (12) (no pagination). doi: 10.1007/s11926-018-0795-4 [DOI] [PubMed] [Google Scholar]
- 63.Min HK, Lee J, Ju JH, et al. Alcohol consumption as a predictor of the progression of spinal structural damage in axial spondyloarthritis: data from the Catholic Axial Spondyloarthritis COhort (CASCO). Arthritis Res Ther 2019;21:187. doi: 10.1186/s13075-019-1970-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.van der Heijde D, Braun J, Deodhar A, et al. Modified stoke ankylosing spondylitis spinal score as an outcome measure to assess the impact of treatment on structural progression in ankylosing spondylitis. Rheumatol Oxf Engl 2019;58:388–400. doi: 10.1093/rheumatology/key128 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Maas F, Arends S, Brouwer E, et al. Incorporating assessment of the cervical facet joints in the modified Stoke ankylosing spondylitis spine score is of additional value in the evaluation of spinal radiographic outcome in ankylosing spondylitis. Arthritis Res Ther 2017;19:77. doi: 10.1186/s13075-017-1285-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66.de Bruin F, de Koning A, van den Berg R, et al. Development of the CT Syndesmophyte Score (CTSS) in patients with ankylosing spondylitis: data from the SIAS cohort. Ann Rheum Dis 2018;77:371–7. doi: 10.1136/annrheumdis-2017-212553 [DOI] [PubMed] [Google Scholar]
- 67.de Koning A, de Bruin F, van den Berg R, et al. Low-dose CT detects more progression of bone formation in comparison to conventional radiography in patients with ankylosing spondylitis: results from the SIAS cohort. Ann Rheum Dis 2018;77:293–9. doi: 10.1136/annrheumdis-2017-211989 [DOI] [PubMed] [Google Scholar]
- 68.Tan S, Yao J, Flynn JA, et al. Quantitative measurement of syndesmophyte volume and height in ankylosing spondylitis using CT. Ann Rheum Dis 2014;73:544–50. doi: 10.1136/annrheumdis-2012-202661 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary File 3. Funnel plot of included studies looking at A) radiographic progression at the spine- TNFi (2 years), B) radiographic progression at the spine- TNFi (≥ 4 years), C) change in the number of syndesmophytes- TNFi, and D) radiographic progression at the spine- NSAIDs.
Supplemental File 1. Search strategy for the systematic review process
Supplementary File 2A) Risk of bias of the included studies i) New-Castle Ottawa scale for observational cohort studies, and ii) Revised Cochrane risk-of-bias tool (RoB 2.0) for randomized controlled trials, 2B) GRADE quality assessment of main outcomes for TNFi and NSAIDs.