POS0029 CHARACTERISING THE CLINICAL AND RADIOLOGICAL PHENOTYPE OF PSORIATIC ARTHRITIS: A CLASS I AND II HLA ALLELES STUDY

Background

Psoriatic arthritis (PsA) is a heterogeneous disease affecting multiple tissues including skin and joints. Few studies have fully explored the association between class I and II HLA alleles and different clinical and radiological manifestations.

Objectives

To examine the association between different class I and II HLA alleles and certain PsA phenotypes.

Methods

Cross-sectional study including patients with PsA over 18 years. All patients fulfilled the CASPAR classification criteria for PsA. Demographic, clinical, laboratory and radiology data were gathered including variables related to joint phenotype, type of skin involvement and response to biological disease-modifying drugs (bDMARDs) treatment. HLA class I (A, B, C) and II (DR, DQ, DP) determinations were performed with high-resolution techniques. We explored possible associations with univariate logistic regression and applied the Bonferroni method to correct for multiple comparisons. bDMARDs resistant patients were defined if failure to two or more bDMARDs.

Results

A total of 246 patients were included. Mean age of the cohort was 59.8 years and 53% were men. Average articular disease duration was 12.4 years. A total of 39.8% patients had axial involvement and 90.7% peripheral, being the polyarticular the most frequent subtype in 48% of cases. Considering the skin, mean disease duration was 22.2 years, being the plaque psoriasis the most frequent subtype in 75.2% of cases. 44.3% of the patients presented a history of dactylitis whilst 24% presented enthesitis. Uveitis was reported in 3.7% of patients and Crohn’s disease in 0.4%. 30.9% of PsA patients had erosions in hand or feet x-ray. More than half (59.3%) were treated with bDMARDs with a median exposure of 2 different drugs.

Table 1 shows the HLA I and II alleles with a frequency greater than 5%.

Table 1.

HLA class I and II alleles with frequencies greater than 5%

GENALLELENFREQUENCY (%)HLA-A02:017923.8701:015115.4129:02298.7603:01267.85HLA-B27:05319.0644:03267.651:01226.4308:01216.14HLA-C07:013911.7806:023811.4804:013410.2712:03329.67HLA-DR07:016018.1301:013711.1803:01288.46HLA-DQ05:016019.802:024916.1703:013912.8702:01299.57

HLA-B*27:05 (OR 2.3, CI 95% 1.01-5.47) and HLA-C*02:02 (OR 2.8, CI 95%, 1.08-7.55) were associated with axial involvement. HLA-DRB*03:01 (OR 3.1, CI 95%, 1.22-8.36) and HLA-DQ*02:01 (OR 3.24, CI 95%, 1.3-8.82) were associated with the polyarticular subtype, HLA-B*08:01 (OR 5, CI 95%, 1.81-14.33), HLA-C*07:01 (OR 2.65, CI 95%, 1.19-5.76) and HLA-DQ*02:01 (OR 2.55, CI 95%, 1.03-6.11) with enthesitis; HLA-A*29:02 (OR 2.68, CI 95%, 1.12-6.75) and HLA-A*11:01 (OR 2.77, CI 95%, 1.06-7.78) with dactylitis; HLA-C*06:02 (OR 2.33, CI 95%, 1.09-5.06) with nail psoriasis; HLA-C*04:01 (OR 3.52, CI 95%, 1.18-9.64) with palmo-plantar psoriasis; HLA-C*06:02 (OR 4.5, CI 95%, 1.33-23.7) and HLA-C*12:03 (OR 3.59, CI 95%, 1.05-19.05) with plaque psoriasis and HLA-A*24:02 (OR 3.03, CI 95%, 1.17-7.92) with erosions. No association was found between alleles and bDMARD resistant PsA patients.

Conclusion

In our cohort, we have confirmed previously described associations in the literature1 and found new significant associations with clinical and radiographic PsA manifestations including class II HLA analysis.

References

[1]Haroon M, Winchester R, Giles JT, et al. Certain class I HLA alleles and haplotypes implicated in susceptibility play a role in determining specific features of the psoriatic arthritis phenotype. Ann Rheum Dis 2016;75:155–162.

Disclosure of Interests

None declared

THU0001 GENOME-WIDE ASSOCIATION STUDY ON JOINT EROSIONS IN RHEUMATOID ARTHRITIS SUPPORTS DIFFERENTIAL PATHOLOGICAL MECHANISMS ACCORDING TO ANTI-CCP STATUS

Background:

Joint damage is the pathological hallmark of rheumatoid arthritis (RA). To identify the genetic variation associated with a higher level of erosions has proven elusive.

Objectives:

The objective of the present study was to perform a genome-wide association study on joint damage in a cohort of RA patients of the Spanish population. Our aims were to provide independent validation of previously reported variants and to identify new candidate risk loci. A stratified analysis was performed based on positivity to ACPA status.

Methods:

A total of 1,135 patients diagnosed with RA using the ACR-EULAR criteria recruited by the IMID Consortium were genotyped using a 550,000 single-nucleotide polymorphism array. Additional SNPs were imputed using the 1KG genome data. Joint damage was performed using the S-score, a simplified radiographic erosion score that has a high correlation with the Sharp-van der Hejde score (1). Association testing of SNPs with joint damage was performed via linear regression with the addition of the years of evolution as covariate. The two main components of genetic variation were also added to adjust for potential population stratification. A total of 50 SNPs representing previously reported loci associated with joint damage were selected. Genetic association was also performed at the pathway level using Pascal.

Results:

45 out of 50 SNPs representing 31 previously reported loci for joint damage could be satisfactorily imputed. Association testing of the whole patient cohort replicated the association withIL2RAandTRAF1. Of relevance, after stratifying for anti-CCP five new loci were replicated:KIF5AandSOSTin ACPA-positive RA andCD40, DKK1andTNFin ACPA-negative RA.IL2RAwas only significant in the ACPA-positive group andTRAF1was not significant in either strata. GWAS on the ACPA-positive cohort and on the ACPA-negative group identified n=7 and n=18 loci with P-values < 1x10-5, respectively. From these, however, only 1 SNP showed nominal significant association in the other patient group. Based on this evidence, we performed a pathway-based analysis to understand the biological mechanisms underlying this difference. Pathway analysis showed 52 biological processes associated with joint damage in ACPA-negative RA and 32 pathways in the ACPA-positive group, with only two shared biological processes between the two groups. Fc Gamma receptor mediated phagocytosis was the topmost biological process associated with erosions specifically in ACPA-negative RA and Signalling by Fibroblast Growth Factor mutants was the top process specific for ACPA-positive patients.

Conclusion:

The results from our study provide suggestive evidence that the genetic basis for joint damage is different according to the presence of ACPA. Replication of the new candidate loci in an independent patient cohort is underway.

References:

[1]Lopez-Lasanta, M., Julià, A., Maymó, J., Fernández-Gutierrez, B., Ureña-Garnica, I., Blanco, F. J., ... & Tornero, J. (2015). Variation at interleukin-6 receptor gene is associated to joint damage in rheumatoid arthritis.Arthritis research & therapy,17(1), 242.

Disclosure of Interests:

Antonio Julià: None declared, Francisco Blanco: None declared, Benjamin Fernandez: None declared, Antonio Gonzalez: None declared, Juan D: None declared, Joan Maymó: None declared, Mercedes Alperi-López: None declared, Alejandro Olive: None declared, Héctor Corominas Speakers bureau: Abbvie, Lilly, Pfizer, Roche, Victor Martinez Taboada: None declared, Isidoro González-Álvaro Grant/research support from: Roche Laboratories, Consultant of: Lilly, Sanofi, Paid instructor for: Lilly, Speakers bureau: Abbvie, MSD, Roche, Lilly, Antonio Fernandez-Nebro: None declared, Alba Erra: None declared, Simon Sánchez Fernandez: None declared, Núria Palau: None declared, Maria Lopez Lasanta: None declared, Adrià Aterido: None declared, Jesús Tornero: None declared, Sara Marsal: None declared

OP0298 STUDY OF RISK OF VERTEBRAL FRACTURES AFTER THE WITHDRAWAL OF DENOSUMAB TREATMENT

Background:

The discontinuation of treatment with denosumab (Dmab) has been associated with a reactivation effect of bone metabolism that manifests itself with a loss of bone mass and an increased risk of vertebral fractures (VF). The incidence and risk factors that may lead to such loss are not clearly established.

Objectives:

Determine the incidence of VF and bone loss in patients who have withdrawn treatment with Dmab and objectify possible associated risk factors.

Methods:

Retrospective review study of patients treated with Dmab and monitored the last two years in monographic osteoporosis consultations. . We selected patients who withdrew treatment with Dmab and registered demographic characteristics, risk factors for osteoporosis and densitometries prior to treatment and during the period of suspension. We identified patients who presented fractures during treatment withdrawal period, assessing: number of fractures, time from withdrawal to fracture presence, location and if they had received osteoactive treatment in that period.

Results:

Of 415 patients treated with Dmab, 83 discontinued treatment. The average age was 63.91 years, 95.2% of them women. The average duration of treatment prior to withdrawal was 2.73 years. 43.4% of the patients had previous fractures, 47.2% vertebral. The data of the previous bone mineral density and during the follow-up are shown in Table 1. 60 patients presented risk factors for osteoporosis, the most frequent being low calcium intake (36.6%) and 15.6% had disease and osteopenizing treatment. 92.7% of the patients had received prior osteoactive treatment. The most frequent reason for withdrawal of Dmab was for therapeutic vacations (56.6%). 39 patients performed post-withdrawal osteoactive treatment, mostly zoledronate (51.3%). During the two years after the rest, 9 patients had fractures (10.9%), seven of vertebral location (77.7%) and ≥ 2 VF were observed in five of them. 5 patients (71.4%) already had fractures prior to the onset of Dmab. The average time from withdrawal from treatment to fracture presentation was 15 months. None of the fractured patients had received treatment after Dmab withdrawal. Although the mean BMD analyzed by DXA at the end of treatment and that the loss of BMD during rest was higher in patients with fracture compared to those without fracture (-7.8% vs -4.3% in the spine and -8.6% vs -4.4% in total femur), the differences were not significant.

Table 1.

BMD values and percentage of BMD change at the start of treatment with denosumab and during two years of withdrawal.

Previous Stop Dmab(n=83)Stop Dmab(n=54)Break Dmab1(n=28)Break Dmab2(n=20)DMO (mean)(g/cm2)T-score(g/cm2)T-score(g/cm2)T-score(g/cm2)T-score Lumbar spine0,861 ± 0,1-2,610,949 ± 0,1-1,940,965 ± 0,2-1,930,920 ± 0,2-2,22 Femoral neck0,735 ± 1-1,960,774 ± 0,1-1,710,758 ± 0,1-1,840,740 ± 0,1-2,02 Total femur0,784 ± 0,1-1,810,823 ± 0,1-1,480,805± 0,1-1,630,801 ± 0,1-1,68% change DMO Lumbar spine12,2 ± 10,3-5 ± 7,9-5,44 ± 7,9 Femoral neck6 ± 8,5-4,2 ± 3,9-5,35 ± 5,8 Total femur3,9 ± 4,2-5,2 ± 4,9-0,33 ± 19,6

Conclusion:

The incidence of VF in patients who interrupted Dmab was 8.43%. Fractured patients had lower BMD gain despite the treatment than non-fractured patients and also the loss of BMD at rest was greater, without significant differences probably due to low number of patients. Neither the presence of previous fractures nor the duration of treatment could be related to the presence of VF at rest.

Disclosure of Interests:

None declared

THU0016 EPIGENOMIC ANALYSIS OF RA PATIENTS SHOWS DISTINCT BIOLOGICAL PROCESSES ASSOCIATED WITH ANTI-TNF RESPONSE

Background:

Blocking Tumor Necrosis Factor (TNF) activity is a successful therapeutic approach for approximately 60% of patients with rheumatoid arthritis (RA). To date, however, the biological basis of the lack of efficacy of anti-TNF agents is unknown.

Objectives:

The objective of present study was to characterize the biological basis of anti-TNF lack of efficacy in RA using an epigenomic data approach in two steps: first, to assess the differential methylation changes between responders and non-responders and second, to use this differential methylation profile in a systems biology approach to infer differential methylated biological modules according to anti-TNF response.

Methods:

A total of n=68 patients diagnosed with RA according to the ACR-EULAR criteria belonging to 16 Hospitals across Spain were recruited. All patients were >18 years old, with more than 6 months of disease evolution and a baseline disease activity of DAS28 > 3.2. Treatment response was defined according to the EULAR criteria at week 12. Good and moderate responders were aggregated into a single responder group. Genomic DNA was collected at baseline and the methylation profile was assessed using the Illumina Infinium EPIC array, which interrogates 850,000 methylation CpG sites across the genome. Differential Methylation analysis, biological pathway association and the systems Biology approach using Protein-Protein Interaction Networks, were conducted using the R statistical language and the Bioconductor libraries.

Results:

From 68 anti-TNF treated patients, n=27 (39.7%) were good responders, n=26 (38.2%) moderate responders and n=15 (22.05%) non-responders at week 12 of treatment. Differential methylation analysis identified two distinctive biological profiles associated with the clinical response: responders were associated to interleukin and cytokine production, and non-responders were associated with biological pathways associated to TGF-Beta production and T cell regulation. Using these differentially methylated profiles, epigenetic modules with differentially methylated hotspots between responders and non-responders were also found. Two epigenetic modules with significant enrichment in inflammatory and interleukin production and immune regulatory processes were validated in an independent patient cohort.

Conclusion:

The epigenetic analysis of whole blood from RA patients using a module-based approach shows reproducible biological mechanisms associated with the response to anti-TNF therapy.

Acknowledgments:

We would like to thank the clinical researchers and patients participating in the IMID Consortium for their collaboration

Disclosure of Interests:

Antonio Julià: None declared, Antonio Gómez: None declared, Antonio Fernández Nebro: None declared, Francisco J. Blanco Grant/research support from: Sanofi-Aventis, Lilly, Bristol MS, Amgen, Pfizer, Abbvie, TRB Chemedica International, Glaxo SmithKline, Archigen Biotech Limited, Novartis, Nichi-iko pharmaceutical Co, Genentech, Jannsen Research & Development, UCB Biopharma, Centrexion Theurapeutics, Celgene, Roche, Regeneron Pharmaceuticals Inc, Biohope, Corbus Pharmaceutical, Tedec Meiji Pharma, Kiniksa Pharmaceuticals, Ltd, Gilead Sciences Inc, Consultant of: Lilly, Bristol MS, Pfizer, Alba Erra: None declared, Simon Sánchez Fernandez: None declared, Jordi Monfort: None declared, Mercedes Alperi-López: None declared, Isidoro González-Álvaro Grant/research support from: Roche Laboratories, Consultant of: Lilly, Sanofi, Paid instructor for: Lilly, Speakers bureau: Abbvie, MSD, Roche, Lilly, Rosario Garcia de Vicuna Grant/research support from: BMS, Lilly, MSD, Novartis, Roche, Consultant of: Abbvie, Biogen, BMS, Celltrion, Gebro, Lilly, Mylan, Pfizer, Sandoz, Sanofi, Paid instructor for: Lilly, Speakers bureau: BMS, Lilly, Pfizer, Sandoz, Sanofi, Raimón Sanmartí Speakers bureau: Abbvie, Eli Lilly, BMS, Roche and Pfizer, Cesar Diaz Torne: None declared, Carlos Marras Fernandez Cid: None declared, Jesús Tornero Molina: None declared, Núria Palau: None declared, Raquel M Lastra: None declared, Jordi Lladós: None declared, Sara Marsal: None declared

SAT0006 SIMULTANEOUS ANALYSIS OF ANTI-CCP, RHEUMATOID FACTOR, ANTI-PAD4 AND ANTI-CARBAMYLATED PROTEIN ANTIBODIES REVEALS INTERACTION EFFECTS WITH RESPONSE TO ANTI-TNF THERAPY IN RHEUMATOID ARTHRITIS

Background:

Blocking of the Tumor Necrosis Factor (TNF) activity is a successful therapeutic approach for 2 out of 3 Rheumatoid Arthritis patients. Identifying the patients that will not respond to this therapeutic approach is a major translational goal in RA. Association of seropositivity to rheumatoid factor (RF) or anti-cyclic-citrullinated antibodies (anti-CCP) with anti-TNF response has proven inconclusive, suggesting that other yet unexplored biomarkers could be more informative for this goal.

Objectives:

We tested the association of two recently introduced biomarkers in RA: anti-carbamylated protein antibodies (anti-CarP) and anti-peptidylarginine deiminase type 4 (anti-PAD4).

Methods:

A prospective cohort of n=80 RA patients starting anti-TNF therapy was recruited and levels for all four autoantibodies -RF, anti-CCP, anti-CarP and anti-PAD4- were measured at baseline. The change in DAS28 score between baseline and week 12 of therapy was used as the clinical endpoint.

Results:

Single marker-analysis showed no significant association with drug response. However, when testing for interactions between autoantibodies, we found highly significant associations with drug response. Anti-CCP and RF showed a positive interaction with the response to anti-TNF therapy (P=0.00068), and anti-PAD4 and antiCarP titers showed a negative interaction with the clinical response at week 12 (P=0.0062). Using an independent retrospective sample (n=199 patients), we validated the interaction between anti-CCP and RF with the clinical response to anti-TNF agents. (P=0.044).

Conclusion:

The results of this study show that interactions between antibodies are important in the response to anti-TNF therapy and suggest potential pathogenic relationships.

Acknowledgments :

We would like to thank the clinical researchers and patients participating in the IMID Consortium for their collaboration

Disclosure of Interests:

Antonio Julià: None declared, Maria Lopez Lasanta: None declared, Francisco Blanco: None declared, Antonio Gómez: None declared, Isabel Haro: None declared, Antonio Juan Mas: None declared, Alba Erra: None declared, Mª Luz García Vivar: None declared, Jordi Monfort: None declared, Simon Sánchez Fernandez: None declared, Isidoro González-Álvaro Grant/research support from: Roche Laboratories, Consultant of: Lilly, Sanofi, Paid instructor for: Lilly, Speakers bureau: Abbvie, MSD, Roche, Lilly, Mercedes Alperi-López: None declared, Raúl Castellanos: None declared, Antonio Fernandez-Nebro: None declared, Cesar Diaz Torne: None declared, Núria Palau: None declared, Raquel M Lastra: None declared, Jordi Lladós: None declared, Raimon Sanmarti: None declared, Sara Marsal: None declared

AB0855 THE “HORSE SADDLE” SIGN: A NEW ULTRASOUND SIGN FOR OSTEOARTHRITIS

Background:

Hand Osteoarthritis is one of the most prevalent rheumatic diseases that can give early visible findings by ultrasound where synovial hypertrophy, effusion, osteophytes and articular cartilage decrease stand out. However, these findings, although sensitive, may not be very specific since they are also observed in inflammatory arthritis. Throughout our clinical practice in osteoarthritis, we have seen, over repeated examinations, an specific morphological change of bone not previously described in the literature. It is a bone extension in the head of the phalanx of the finger joints that causes a deformity that we have called “horse saddle” and that is typically located in the proximal and distal interphalangeal joints of the fingers. This sign can be seen in the longitudinal exploration of the palm of the hand by grayscale ultrasound. This sign that we have not found specifically described in the literature reviewed to date is considered to be useful for the diagnosis of osteoarthritis.

Objectives:

To Assess the sensitivity and specificity of the “horse saddle” sign in the diagnosis of osteoarthritis.

Methods:

An exploratory clinical comparative cross-sectional study where an ultrasound of the hands and comparative radiographs in PA view were performed on patients with osteoarthritis, inflammatory arthritis and healthy patients seen in the Rheumatology clinic of the Vall de Hebron Hospital. Age, sex and time of evolution of the disease were collected as clinical variables. The MCP, PIP and DIP joints from the second to the fifth finger of both hands were viewed with grayscale in longitudinal and transverse plane of both the dorsal and palmar face, assessing for osteophytes, synovitis and the horse saddle sign. A General Electric Logiq S8 machine was used with an 8-13 MHz linear probe. All patients signed an informed consent and approval was obtained from the hospital ethics committee. The statistical analysis was carried out with Stata 15.1.

Results:

A total of 38 patients with osteoarthritis, 20 patients with inflammatory arthritis (8 psoriatic, 9 RA, 1 LES, 1 PMR and 1 Sjögren) and 2 healthy patients were assessed. It was found that the horse saddle sign had a sensitivity of 66.7% and specificity of 86.4% in osteoarthritis showing a p-value of 0.052 by means of the chi-square test. 87% of patients with the horse saddle sign had osteoarthritis and only in 2 patients with RA and in the patient with LES. In contrast for osteophytes a sensitivity of 100% was observed with a specificity of 45.45% (p of 0.001) and for synovitis a sensitivity of 53.3% and specificity of 77.27% was observed with p of 0.039.

Conclusion:

The horse saddle sign is an ultrasound sign with good sensitivity and specificity for the diagnosis of hand osteoarthritis comparable to other classic ultrasound signs such as osteophytes and synovitis.

References:

[1]Iagnococo A, Palombi G, Valesini G. Role of ultrasound in osteoarthritis. Rev Esp Reumatol. 2001; 28: 301-6.

[2]Abraham AM, Pearce MS, Mann KD, Francis RM, Birell F. Population prevalence of ultrasound features of osteoarthritis in the hand, knee and hip at age 63 years: the Newcastle thousand families birth cohort. BMC Musculoskelet Disord. 2014; 15: 162. DOI: 10.1186/1471-2474-15-162.

Disclosure of Interests:

None declared

Comparison of total, free and bioavailable 25-OH vitamin D determinations to evaluate its biological activity in healthy adults: the LabOscat study

Determination of different forms of 25-OHD (total, free and bioavailable) in healthy young women does not offer additional advantages over standard 25-OHD_T for evaluating vitamin D deficiency. In these subjects 25-OHD_T values <15 ng/ml would be more appropriate for defining this deficiency.

INTRODUCTION

Determination of 25-OH vitamin D serum levels (25-OHD) constitutes the method of choice for evaluating vitamin D deficiency. However, vitamin D-binding protein (DBP) may modulate its bioavailability thereby affecting correct evaluation of 25-OHD status. We analysed the impact of the determination of 25-OHD (total, free and bioavailable) on the evaluation its biologic activity (estimated by serum PTH determination) in healthy young women.

METHODS

173 premenopausal women (aged 35-45 yrs.) were included. We analysed serum values of total 25-OHD (25-OHD_T), DBP, albumin, PTH and bone formation (PINP,OC) and resorption (NTx,CTx) markers. Free(25-OHD_F) and bioavailable (25-OHD_B) serum 25-OHD levels were estimated by DBP and albumin determinations and also directly by ELISA (25-OHD_F-2). We analysed threshold PTH values for the different forms of 25-OHD and the correlations and differences according to 25-OHD_T levels <20 ng/ml.

RESULTS

62% of subjects had 25-OHD values <20 ng/ml and also had significantly lower 25-OHD_F and 25-OHD_B values, with no significant differences in bone markers and PTH values. The PTH threshold value was similar for all forms of 25-OHD (∼70 pg/ml). Women with PTH values >70 had lower 25-OHD_T (15.4 ± 1.4 vs. 18.3 ± 2.7, p < 0.05) and 25OHD_B values (1.7 ± 0.2 vs. 2.2 ± 0.09, p < 0.05). The different forms of 25OHD were significantly intercorrelated, with marginal correlations between PTH and 25-OHD_T (r = -0.136, p = 0.082).

CONCLUSIONS

Determination of different forms of 25-OHD in healthy young women does not offer additional advantages over standard 25-OHD_T for evaluating vitamin D deficiency. In these subjects 25-OHD_T values <15 ng/ml would be more appropriate for defining this deficiency.

Adsorptive granulocyte/monocyte apheresis for the treatment of refractory rheumatoid arthritis: an open pilot multicentre trial

OBJECTIVE

To assess the efficacy and safety of adsorptive granulocyte and monocyte apheresis (GCAP) in patients with refractory rheumatoid arthritis (RA).

METHODS

Patients with active and refractory RA were treated with weekly GCAP sessions using a column filled with acetate beads (Adacolumn) over five consecutive weeks. Clinical assessments and response to therapy were analysed at weeks 5, 7, 12 and 20 in an open multicentre trial. The primary outcome measure of clinical response was 20% improvement in the American College of Rheumatology criteria (ACR20) at week 20. EULAR (European League Against Rheumatism) response criteria, based on the disease activity score for 28 joints (DAS28) and disability using the Health Assessment Questionnaire (HAQ), were also assessed.

RESULTS

Of 27 patients, 81.5% were women with mean disease duration of 14.4 yr. The mean number of previous disease-modifying antirheumatic drugs (DMARDs) was 3.7, and 48.1% of patients had previously failed on biologicals. On an intention-to-treat basis, 40.7% of patients achieved an ACR20 and 44.4% a therapeutic EULAR response at week 20. These percentages were 50 and 54.5% in 22 patients who completed the trial. In the 10 completers who had previously failed on biologicals, an ACR response was achieved in four patients (ACR20, two; ACR50, one; ACR70, one). A significant decrease was recorded in different ACR response components, including the tender joint and swollen joint counts, pain score and patient and physician global disease assessments, as well as the DAS28 index; most of them improved after week 5. ESR and CRP, but not the HAQ score, had decreased significantly at week 20. The treatment was well tolerated and only one serious adverse event related to the study procedure was documented (sepsis due to a catheter infection).

CONCLUSIONS

GCAP treatment led to significant clinical improvement in a subset of patients with RA who had failed to respond to DMARDs or biologicals. Further large, placebo-controlled studies are warranted to fully assess the therapeutic value of GCAP for refractory RA.