The 11th International Congress on Spondyloarthritides was held in the beautiful city of Gent in Belgium, at the Opera House between the 4 and 6thof October 2018. The Congress is held every 2 years, and particularly focusses on axial spondyloarthritis (axSpA) for international specialists in this field, along with scientists, allied health professionals and patients. A particular emphasis this year was on basic and translational research, genetics and imaging.
The conference was opened by Helena Marzo-Ortega (co-President) who outlined the 20 year history of the Congress. She also introduced the first of three basic science lectures, based around the role of the gut in inflammation and the microbiome.
First, Tom Cupedo (The Netherlands, INV2) described how innate lymphoid cells are the cellular regulators of intestinal regeneration after tissue damage, and ensure that intestinal stem cells persist, particularly through the action of ILC3 derived IL22. This particular cell line, and its derived cytokines, were proposed as a possible target for harnessing stem cell regeneration when normal homeostasis is lost.
Later, Robert Colbert (USA, INV4 ) outlined the relationship between the microbiome and gut inflammation in HLA B27 transgenic rats. Using Fischer and Lewis HLA B27 positive rats, compared to wild type, distinct differences in the expression of multiple cytokines from ceacal samples were found, but with overlap between Fischer and Lewis. Pathways that were over expressed included TNF, IL17/12/23, osteoclast differentiation, TCR and adaptive immune responses. Pathways under expressed included metabolism of lipids, oxidation, fat and vitamin absorption. HLA B27 associated microbial differences were largely non-overlapping in Lewis and Fischer rats though (Gill T et al, A&R 2018;70:555), corresponding to those with and without arthritis.
The next session comprised of three hugely interesting presentations regarding gender differences in axSpA. Lianne Gensler (USA, INV5) started by exploring gender differences in the clinical phenotype and response to treatment in axSpA. This excellent overview, commenced with highlighting the differences between gender and sex, and ensuring we use the correct terminology. She suggested the possibility of sex dimorphism of the pelvis and lumbar spine as a contributor to differences in instability, mechanical stress or disc disease between the sexes. Women with axSpA on average have a 7 month longer delay to diagnosis than men, and a 2 fold delay in diagnosis in women who also report widespread pain. Women experience more dactylitis, arthritis, enthesitis. Furthermore, women report more pain, have lower pain tolerance, poorer sleep and higher responses to all elements of the BASDAI outcome measure, but often these outcomes in studies are confounded by gender behaviour differences (i.e. women are ‘expected’ to tolerate these symptoms differently). In addition, differences exist between sexes of the intra dermal pain neural density. Regarding female axSpA patients and their inclusion in studies, they are generally less likely to be prescribed anti TNF or other treatment, have lower drug persistence, higher treatment switch rates and are more likely to withdraw from randomised controlled trials. Overall, they are less likely to reach study endpoints. They may also be less likely to be recruited to studies and often studies are not powered sufficiently to evaluate female response, and we must consider sex as an effect modifier rather than just a confounder.
Eric Gracey (Canada, INV6) outlined the biological differences underpinning some of these differences. He presented differences both in the innate and adaptive immune systems between sexes throughout the lifecycle, such as males having greater inflammatory response, higher levels of natural killer cells through child and adulthood, higher IgE in utero and IgA/M and T reg cells in childhood, and higher CD 8 T cells from puberty onwards. Females have higher immunoglobulins, B cells, and T cell proliferation from puberty onwards (Nat Rev Imm 2016; 16(10):626-38). Discrepancies in IFNgamma levels and Th1 cells were shown between men and women in HC and AS (A&R, 2016;68(3):679-89), and higher IL17 is seen in men with AS. Interestingly, it was noted that in most animal models, the sex of the animal is not reported, but generally male animals will be used.
Walter Magerl (Germany, INV7) outlined the fundamental difference in all types of pain sensation (pressure, cold, sharp, mechanical) in women only from puberty onwards. Women have been found to be 6-10% less physically active than men (Hands et al. J Women’s Health, Issues and Care, 2016;5:4) with poorer sleep and higher stress.
The day was concluded with the Inaugural Gent SpA Oration, to honour a clinical or scientist with an established career trajectory in the field. This was given by Joachim Sieper (Germany, INV8) who gave a comprehensive summary of his career and research to date.
The ‘auto inflammation axis’ session during Day 2,commenced with John Forrester (UK, INV12) discussing eye disease in SpA. He summarised the non-HLA polymorphisms known to be associated with uveitis (FOXO1, TRAF5, ERAP1, IL23, IL18R1). He proposed that eye disease in SpA may be an auto inflammatory phenomenon rather than auto immune. The evidence for this came from multiple sources, including Blau’s syndrome (Saren et al. 2017) and other auto inflammatory conditions (Muckel-Wells, CINCA), altered dendritic and T cell populations in AS (Wright et al. 2016) and previously demonstrated high IL1 levels in patients with acute anterior uveitis (Chen et al. 2015), although clinical evidence for this mechanism is weak. In addition, he proposed that uveitis may in some patients be due to latent infection such as with cytomegalovirus, causing stress of the endoplasmic reticulum in the presence of HLA B27 (Robinson et al. 2016, Nguyen et al. 2014, Kim et al. 2011), and precipitation of an SpA-inflammatory bowel disease-genitourinary-skin-microbiome syndrome through dysfunctional autoinflammatory immune regulation.
The increasing specificity and efficacy of treatments for skin and joints was presented by Diamant Thaçi (Germany, INV13).Initially, treatment with anti IL17 agent Secukinumab was discussed and evidence was shown of persistent remission of skin disease (defined by the PASI, <2.7) for more than 1 year after treatment was withdrawn. Preliminary evidence was also demonstrated for Risankizumab (anti IL23, pooled 90 and 180mg doses) in PASI90 (80% patients achieving this response threshold) with persistent response out to week 48 of treatment (46.3% 90mg, 47.6% 180mg vs 10% Ustekinumab) (Papp KA et al. NEJM 2017 20;176(16):1551-1560). Median time to loss of PASI90 response was 309 and 345 days in Risankizumab 90 and 180mg, vs 112 days for Ustekinumab.
This session was followed by a hugely stimulating and exciting debate between Atul Deodhar (USA, INV14, for motion) and Pedro Machado (UK, INV15, against motion), stating their opinions on whether there is no place for ‘Treat to Target’ (T2T) in the treatment of axSpA. Particular points of debate included the concern for Atul Deodhar that T2T may limit the individuality of treatment planning for patients (reducing the capacity for personalised medicine) and whether this really differed from routine care. Whilst Pedro Machado clearly demonstrated a strong evidence base for optimised disease control on multiple disease related outcomes, and the 2017 international task force statement on T2T in axSpA (Smolen JS et al. ARD 2018;77(1):3-17). Further discussion points outlined the concern regarding the chosen outcome measure threshold for T2T, and how this management strategy may be instituted in already stretched healthcare systems.
Later during day 2, specific conference abstracts had been selected for oral presentation; 6 in total during this session. It was during this period that I presented my oral abstract.
- O1: Eva Klingberg (Sweden)– Gut dysbiosis in AS is associated with increased faecal calprotectin
- O2: Francesco Ciccia (Italy)– Inflammasomes activation occurs in the inflamed tissues of AS patients and rives IL23 expression and ILC3 expansion
- O3: Leonie van Duivenvoorde (The Netherlands) – Transmembrane TNF signalling through TNFRI induces SpA-like inflammation, whereas signalling through TNFRII is cruicial for new bone formation
- O4: Elizabeth Reilly (UK)– Prognostic markers in axial spondyloarthritis – alpha 1 anti trypsin and beta 2 microglobulin may differentiate AS from nr-axSpA, mechanical back pain and healthy controls
- O5: Daniel Wendling (France)– Impact of gut involvement in early SpA – the DESIR cohort
- O6: Sophie Hoekstra (The Netherlands)– Gender differences in TNFi treatment adherence and response in AS patients: A prospective longitudinal cohort study
The day was concluded with sessions on genetics and genomics in SpA.
Matthew Brown (Australia, INV19) commenced by outlining the GWAS project on ~1000 subjects. As a result of lack of IL17A found in AS compared to IBD or psoriasis without response to anti IL23p19, he hypothesised whether IL17 in AS, is produced mainly by upregulation of IL23R signalling rather than IL23 itself. Given that modest evidence exists that offspring of affected mothers are more likely to develop AS than from affected fathers, he proposed that females may have a higher threshold for the development of AS, with less genetic load required for AS. Diagnostic bias however must be appreciated in studies concerning prevalence across the sexes. In HLA B27 negative patients, previous studies have suggested increased frequency of MEFV genetic variants (Cosan et al. A&R 2011), which was replicated in this GWAS study, with high serum IL1 levels in specifically the MEFV694V carriers.
Donal O’Shea (Ireland, INV21) then presented ‘Operation transformation: the impact of genes and life style in joint disease’. This insightful session described the impact of obesity on immune cells and genes, and the required steps to address this challenge. Physical activity has a direct effect on reduced mortality (up to a point), but further influences on body weight include the increasing external pressures from society, activity environment, food consumption and production and individual psychology. Obesity has been seen to effect invariant natural killer cells (known as the ‘swiss army knife’ of the immune system), a key cell in the innate immune system interacting with multiple other cells and determining the downstream responses. Further effects are seen on mucosal associated invariant T cells, which are increased in childhood obesity but decreased in adult obesity, producing less IL10 and more IL17. He suggested that obesity needs to be recognised as a modifiable driver in all disease, across all socioeconomic groups, with appropriate education delivered at all ages, including the involvement from primary care as the main site for the identification and education of overweight patients, with support from bariatric secondary care. It is essential this is supported by appropriate changes in advertising, sponsorship, food labelling and healthy lifestyle promotion.
Day 3 commenced with a series of short oral presentations. The focus for the first part of this session was upon imaging in SpA, with 2 initial presentations from Xenafon Baraliakos (Germany, O7 & O8). In his first presentation, he discussed his research examining the cellular composition of fatty lesions on MRI in vertebral bodies of patients both with AS and degenerative disc disease (DDD), using decalcified biopsies. Differing compositions of fat, inflammation and fibrosis were seen between biopsies of AS and DDD, with a predominance of adipocytes in AS. During his second presentation, results from an assessment of the prevalence of axial MRI bone marrow oedema (BMO) and fat metaplasia suggestive of axSpA in healthy participants <45 years of age from a large community study were discussed. BMO was noted in 18% of the cohort, with 44% demonstrating more than 5 such lesions, most commonly in the thoracic spine. Similarly, fatty lesions were most commonly seen in the thoracic spine , with 39.1% of those with these changes having more than 5 lesions.
Results from a further imaging study, ECHOSPA, were then presented by Walter Maksymowych (Canada, O9).MRIs were read by 2 experienced, independent readers, and typical active and structural axSpA lesions recorded, to validate the ASAS MRI lesion definitions in an axSpA cohort. Both inflammatory and structural lesions were seen in early stage axSpA, with demonstration of reliable detection of SIJ lesions, including backfill using the ASAS definitions. It was suggested a SPARCC BMO score of >=3 and erosion score of >=2 may optimally reflect active and structural lesions typical of axSpA.
Alexandre Sepriano (The Netherlands, O10) presented data from the DESIR cohort, from patients with inflammatory MRI features and the possible effect on structural damage seen at 5 years. Patients had disease of less than 3 years duration, and each MRI at baseline and 5 years was read by 3 independent readers. ASAS definitions for inflammatory change were used, and >=3 fatty lesions or erosions were deemed to be evidence of structural damage at SIJ and spine. Multivariate analysis included age, sex, HLA B27, smoking, ASDAS and NSAID/TNFi treatment as confounders. They found that patients with baseline inflammatory lesions have a higher probability of net structural fatty lesions compared to those negative for inflammatory lesions at the SIJ and spine, regardless of CRP. Increased odds ratios were also seen for baseline BMO and erosions at the follow at the SIJ.
Desiree van der Heijde (The Netherlands, O11) reported results for Bimekizumab phase 2b study (dual neutralisation of IL17A and IL17F) in active AS at 12 weeks (BE AGILE study). Given the 50% structural homology of IL17 A and F and overlap in biological functions, it is hypothesised that the effect would be larger with dual neutralisation than with partial IL17A blockade only. 303 patients with BASDAI >4 and spinal pain >4 (and maximum 1 previous TNFi) were randomised to placebo, 16, 64, 160 or 320mg Q4W. At week 12, they were reallocated to either 160 or 320mg Q4W. 42.6%, 46.7% and 45.9% of patients receiving 64, 160 and 320mg Bimekizumab respectively reached ASAS40 response by week 12, compared to only 13.3% of the placebo group. Divergence was seen between groups as early as week 1. Similar dose responsive results were seen in ASAS5/6 and ASDAS. The commonest side effects seen were nasopharyngitis and headache, with oral candidiasis in 1.2%. 3 patients had severe TEAEs and there was 1 death in the treatment arm (cardiac arrest considered not related).
Philip Carron (Belgium, O12) summarised results from their study examining a high need for TNFi after withdrawal in patients in remission from early peripheral SpA. Exclusion criteria included prior DMARDs, TNFi or corticosteroid treatment although concomitant NSAIDs were allowed, and patient had a short (<12 weeks) symptom duration. Patients on TNFi (Golimumab) were required to be in clinical remission between weeks 24 and 48, and then treatment was withdrawn. If they relapsed, patients were restarted on Golimumab up to week 104, with a tapering strategy of Methotrexate added at this point and Golimumab stopped at week 116 if in remission. Of the initial 60 patients included, 49 reached sustained clinical remission and stopped TNFi. 47% went on to relapse, with a median time to relapse of 32 weeks. Further, 22 patients entered the taper strategy commencing Methotrexate, 77% of which required Golimumab to be re-instated. All of these had peripheral arthritis, 76% male, 59% psoriasis, 47% dactylitis, and 41% HLA B27 positive.
Salima van Weekly (The Netherlands, O13) concluded the session, presenting the Dutch recommendations for physical therapy in axSpA. These had been developed as there has been significant variation in delivery of physiotherapy in Holland, without specific axSpA guidelines. Particular areas in the new recommendations included appreciated the comorbidities of patients with axSpA such as increased risk of falling and fractures, and the limitations on pulmonary functioning. Manipulations are contra-indicated, and education needs to be provided regarding activities which may be associated with increased fractures risk such as for patients with long disease duration, severe ankylosis and low bone mass.
During the session ‘bone and SpA’ on day 3, Maurizio Pacifici (USA, INV22 )outlined the role of BMP signalling in pathologic ossification and angiogenesis, through the example of multiple osteochondroma, an autosomal dominant disorder of 1 in 50,000. In this condition, bony /cartilage tumours form within the perichondrium flanking growth plates in juvenile and adolescent patients. This process commences due to an inciting event, causing recruitment of progenitor cells, cartilage differentiation, maturation and hypertrophy, leading to endochondral bone formation, through pro chondrogenic BMP signalling and reduced anit-chondrogenic signalling by FGF. This mechanism therefore could be a possible target for future therapy in conditions where ossification, such as AS, are seen. Indeed, Clementia Pharmaceuticals are currently pursuing iindirect BMP signal inhibition with Palovarotene in phase 2 trials in multiple osteochondroma and phase 3 trials in fibrodysplasia ossificans progressive.
George Schett (Germany, INV23 )followed, presenting ‘can diet affect arthritis and inhibit bone loss?’. He explained that historic, stone age diets were rich in fibre, including lots of fruits and nuts foraged from the habit. These fibres are fermented by the microbiota to short chain fatty acids (SCFA); acetate, proprionate and butyrate. He presented evidence by Lucas et al. of increased bone mass in high fibre diets with reduced osteoclast numbers. SCFA have also been demonstrated to inhibit postmenopausal bone loss and induce metabolic reprogramming in osteoclasts with induction of glycolysis. North et al. (Eur J Clin Nutr, 2009, 63;921-33) also reviewed 7 studies in a meta analysis of the effect of dietary fibre on systemic inflammation, over at least a 2 weeks period in adults. Six of these demonstrated a reduction in CRP of between 25-54%, with one study in obese individuals failing to show any effect.
Dirk Elewaut (Belgium, INV24 )then concluded this section, with evidence of the effect of mechanical stress on SpA. He started by introducing the idea of SpA as a primary entheseal disease, with secondary synovial involvement, more in keeping with PsA than RA. Greater dependence is seen on PGE2 and Il17/23 than in RA, with less involvement of IL6 and less effect from treatments such as Methotrexate. However, previous work has demonstrated an improvement in active RA with paralysis, the development in manual workers, and similar picture of greater disease progression in SpA in those with physically demanding jobs. Through mouse models (using tail suspension, removing mechanical stress from the hindlegs, compared to forefeet and free running mice), his study has demonstrated greater enthesitis in the forefeet and free running mice compared to those without mechanical stress. It was proposed that mechanical stress could be the switch from the autoimmune to inflammatory phases of disease in SpA, and may be responsible for the pattern of joint involvement. This would have significant potential implications for physical therapy in SpA however.
Two particularly clinically relevant sessions drew the congress to a close: Desiree van der Heijde (The Netherlands, INV26) and Xenafon Baraliakos (Germany, INV27). In the first of these presentations, Desiree outlined the potential use of low radiation CT (4mSv, compared to xray cervical spine 2.6mSv) of the spine to assess structural progression in SpA. The European Commission state that radiation dose ranges of 1-10mSv in adults denotes a risk of 1 in 10,000, and the degree of benefit to society should be deemed ‘moderate’ and ‘aimed directly at the diagnosis, cure or prevention of disease’, with a stated background radiation level of 2.6mSv per year in the Netherlands. Tan et al. (ARD, 2015, 74;437-43) have previously reported greater sensitivity to change over 1 year using CT than for mSASSS or ASspiMRI-c. The study was carried out in 2 European centres, over 2 years, and patients were required to have at between 1 and 18 cervical or lumbar syndesmophyte on xray and no active inflammation on MRI. 50 patients were included (84% male, mean age 50 years, 86% HLA B27, 38% raised CRP, median syndesmophytes at inclusion 4.5). Each vertebral unit (lower vertebral body, intervertebral disc, upper vertebral body) were assigned 2 scores for each anterior corner, and divided into 4 quadrants. They were scored 0 – absent to 3 – bridging syndesmophytes /ankylosis. Results showed that most abnormalities were located in the thoracic spine, with 50% having ankylosis at this spinal segment. Using the mSASSS, a mean 2.4 point change was seen over 2 years (SD 3.8) whilst using CT scoring, a mean 17.9 point change (SD 22.1) was seen over the same time period. Using CT, there was a greater sensitivity to change (resulting in fewer patients required to power studies), with the most frequent area for syndesmophytes was in the thoracic spine, which is not assessed using mSASSS and poorly imaged using conventional radiographs.
Xenafon Baraliakos then concluded by describing the current evidence for measuring bone proliferation in axSpA. Fat metaplasia at SIJ on MRI increases the propensity for disease progression in the spine in SpA (Maksymowych W et al. RMD Open 2017). From Haroon N et al. (A&R 2013) early treatment with biologics has been shown to be associated with less radiographic progression, with further evidence using Certolizumab Pegol from van der Heidje et al. (ARD 2018) and reduced disease activity from Molnar C et al. (ARD 2017).
The congress was closed with presentations for best oral abstracts, and thanks to all those who took part and those that attended this hugely successful and engaging meeting.
The SpA teams from Norwich and Bath enjoying some relaxing time in Gent.
