Can Rheumatoid factor and Anti-cyclic citrullinated peptides predict true remission in rheumatoid arthritis patients? Study of relationship between autoantibodies levels and muskuloskelatal ultrasound ndings in a sample cohort of Egyptian patients in clinical remission

Background True remission is the ultimate goal for rheumatoid arthritis (RA) therapy. Our aim was to investigate the relationship between serum levels of rheumatoid factor (RF) and anti-cyclic citrullinated peptides (Anti-CCP) and ultrasonographic (US) ndings in Egyptian RA patients in clinical remission. Methods Using data from a cross-sectional study on 50 RA patients in clinical remission or low disease activity (LDA) as dened by disease activity score (DAS28-ESR) cutoff points, performed in Alexandria University Hospital; we analyzed statistical relationships and correlations between RF, Anti-CCP) and Gray Scale (GS) and Power Doppler (PD) US using US7 score. US remission was dened as on a GS ≤ 1 and PD = 0. were analyzed using IBM SPSS software package version 20.0. (26) (Armonk, NY: IBM Corp). Qualitative data were described using number and percent. Quantitative data were described using range (minimum and maximum), mean, standard deviation and median. Mann Whitney (U) test for abnormally distributed quantitative variables, to compare between two studied groups. Spearman coecient (rs) to correlate between two distributed abnormally quantitative variables. The signicance of the obtained results was judged at the 5% level. remission


Background
Achieving rheumatoid arthritis (RA) disease remission is becoming an imperative goal presently, thanks to the revolutionary treatment protocols that have been introduced to the disease management in the past twenty years (1,2) . Although reaching point remission is a pivotal step in judging treatment success, sustained disease remission will remain the ultimate outcome that both patients and clinicians long to achieve. (3) Disease remission has been described in many aspects; including clinical, imaging and immunologic remission. Clinical remission depends on the use of composite indices that include joint counts, acute phase reactants and pain assessment measures. These indices can be stringent such as simpli ed disease activity index (SDAI), or less stringent in assessing disease remission as disease activity score (DAS28). (4,5) However, they were proved to be lacking sensitivity to detect low levels of in ammation in many studies. (6) Musculoskeletal ultrasonography (MSUS) has been demonstrated to detect subclinical synovitis not appreciated by routine clinical or laboratory assessments. (7) Moreover, it was recently suggested that they have the ability to predict the duration of disease remission and the occurrence of ares. (8,9) Beside the pathogenic and diagnostic functions of autoantibodies in RA patients, rheumatoid factor (RF) and anti-cyclic citrullinated peptides (Anti-CCP) have a clear association with the disease in ammatory process and the progression of structural damage, which emphasizes their role in the disease prognosis. (10,11) Serum levels of RF and Anti-CCP may even change during treatment in some patients who have good clinical response. (12) Although patients may continue having high concentrations of auto antibodies, studies proved that this does not prevent the occurrence of clinical remission. Yet, Anti-CCP positive patients seemed to face more ares. (10,13) Recently, many studies proved the presence of subclinical synovitis-as detected by MSUS-in the joints of RA patients in clinical remission regardless the tightness of the used indices. Little evidence is present to analyze the association between autoantibodies and MSUS patterns in joints of RA patients who achieved clinical remission.

Methods
We performed secondary analysis from a cross-sectional study (14) that included 50 Egyptian patients; 37 (74%) females and 13 (26%) males) with RA, diagnosed according the ACR/EULAR 2010 criteria for diagnosis of RA (15) , and ful lled the cut-off values of clinical remission or low disease activity (LDA) according to DAS28-ESR score. (16,17) The patients were recruited from those attending the Outpatient Clinic of Physical Medicine, Rheumatology and Rehabilitation Department and the Rheumatology and Immunology unit, Internal Medicine department, Faculty of Medicine, Alexandria University, Patients with osteoarthritis, any systemic disease with in ammatory arthropathies and hepatitis C virus (HCV) arthritis were excluded. An informed consent was given by each patient and the source study was approved by the ethics committee of the Faculty of Medicine. (14) Clinical assessment All patients were subjected to full history taking, including: disease duration, duration of clinical remission or low disease activity and full drug history. The patient's general health was evaluated using a Visual Analog Scale (VAS) of 100 mm, in addition to complete physical examination including 28-tender joint count (28-TJC) and 28 swollen joint count (28-SJC). Disease activity was then assessed using DAS28-ESR score. (16,17) Clinical remission was de ned as a DAS28-ESR score of < 2.6, while LDA as a DAS28-ESR score of ≥ 2.6 and ≤ 3.2. (17)

Patients characteristics
We analysed the data of 50 RA patients, 37 (74%) of whom were females and 13 (26%) were males. Their mean age was 49.58 ± 9.14 years, disease duration mean was 6.75 ± 3.82 years.

Musculoskeletal ultrasonographic examination
The Gray Scale (GS) and Power Doppler (PD) ultrasonography (US) examination was performed using high frequency broadband linear array transducer at 10-18 MHz. PD settings were optimized to enhance the sensitivity for detecting synovial vessels without or with minimal artifact (20) .
The US examination was performed in two perpendicular planes, according to EULAR guidelines (21) . We followed the OMERACT standardized de nitions of US pathological ndings (22) , and scored according to US7 score (23) The US remission was de ned as on a GS ≤ 1 and PD = 0 (24, 25) . While those patients with GSUS > 1 and/or PDUS ≥ 1 were considered to have ultrasonography activity.
Qualitative data were described using number and percent. Quantitative data were described using range (minimum and maximum), mean, standard deviation and median. Mann Whitney (U) test for abnormally distributed quantitative variables, to compare between two studied groups. Spearman coe cient (rs) to correlate between two distributed abnormally quantitative variables. The signi cance of the obtained results was judged at the 5% level.
Patients had a mean ESR of 12.48 ± 4.15 and mean CRP of 2.65 ± 1.46. 7 (14%) of the patients were RF negative patients and 43(86%) RF positive patients. RF value ranged from 10.6-678 IU/ml, with a median of 45.95 IU/ml. The Anti-CCP of the studied patients ranged from 2.8-2940 U/ml, with a median of 36.5 U/ml. There were 21 (42%) Anti-CCP negative patients, and 29 (58%) Anti-CCP positive patients.
Thirty-four (68%) patients were in clinical remission and 16 (32%) were in low disease activity according to cutoff values of DAS28-ESR score, which had a mean of 2.52 ± 0.39 (table 1).

MSUS characteristics
As examined by the US7 score, 21 (42%) patients were found to be in ultrasonographic (US) remission, while 29 patients (58 %) were found to have subclinical ultrasonographic activity. All patients in LDA showed subclinical US activity, while 61.8% (21) of patients in clinical remission were in ultrasonographic remission, and 38.2% (13) of patients in clinical remission had a degree of subclinical ultrasonographic activity (table 2).
To test the relationship between the ultrasonographic ndings and the autoantibodies levels, the comparison between patients in US remission and those with activity yielded a statistically signi cant difference for both RF (p = 0.03) and Anti-CCP (p = 0.021); where patients with US activity had higher autoantibodies levels (table 3). Further comparison revealed a statistically signi cant difference between patients in clinical remission with US remission and those in clinical remission with US subclinical activity; regarding Anti-CCP only (p = 0.006), but not RF (p = 0.086) (table 4). Patients in clinical remission with subclinical US activity had higher Anti-CCP. This latter group of patients included 7(53.8%) males and 6 (46.2%) females, only 3 (23.1%) of the patients had RF negative and 10 (76.9%) were RF positive, 8 (61.5%) of which had highly positive RF (> 3 times upper limit normal-ULN). The Anti-CCP in this group was positive in 11 (84.6%) and negative in 2 (15.4%), 9 (69.2%) patients had highly positive Anti-CCP (> 3 times ULN). In addition, in the same group 7 (53.8%) were treated with cDMARDS and 6 (46.2%) with bDMARDS (table 5).
Based on the previous results, we examined the correlation between Anti-CCP and US7 score results in the aforementioned group, this proved a strong positive correlation between Anti-CCP and synovitis score by power Doppler US (PDUS) (rs = 0.553, p = 0.001), and a similar positive correlation between Anti-CCP and tenosynovitis/paratenonitis score by gray scale US(GSUS) (rs = 0.389, p = 0.023); while none of the US7 score components correlated with Anti-CCP in patients in true remission (table 6).

Conclusion
Our study demonstrated that patients in clinical RA remission with subclinical US activity had higher serum levels of Anti-CCP. Since both have been proved to be related to joint damage progression; Such an association should guide the change treatment options while managing those patients, in order to decrease the incidence of relapse and offer an optimum disease outcome.

Discussion
The occurrence of disease relapse is not an infrequent event in RA patients, even in those with sustained or prolonged periods of remission. Our cross-sectional study demonstrated the association between serum levels of auto antibodies, RF and Anti-CCP, and the presence of subclinical US signals in the joints of RA patients classi ed as being in clinical remission or LDA. Previous studies proved Anti-CCP to be a strong predictor of radiographic progression in RA patients in remission (27) . Similarly, residual PDUS activity in the clinically inactive joints can predict joint damage and the occurrence of erosions. On the other hand, Boetres at al. who investigated the seroreversion rates in patients with long-standing DMARD-free status, found that antibodies disappearance is not a frequent occurrence in such patients and stated that "this form of immunological remission should therefore not be a treatment target". (32) However, they depended only on the clinical disappearance of signs and symptoms of synovitis, which is well known to be lacking in describing true remission and missing the advantage of US of detecting the subclinical synovitis that may be a clue to disease relapse that happened to a number of their patients.
We believe that if the ultimate goal of treat to target therapy is actually a longstanding drug-free "true" remission, then the association between high levels of serum antibodies and US detected subclinical synovitis should be put in consideration while designing treatment plans for RA patients even during remission.
A limiting point in our study was its cross-sectional nature, that prevented us from studying the association of serum antibodies and US detected subclinical synovitis on a longterm during remission, therefor longitudinal studies are mandatory in this respect. Another caveat is the small number of participants that constituted a statistical challenge, this is attributed to the fact that this was a singlecenter study.    Qualitative data were described using number and percentage.
Quantitative data was expressed using Min. -Max. , Mean ± Sd. and median.