Article Text
PDF
Abstract
Objective Rheumatoid arthritis (RA) is characterised by the presence of autoantibodies, among which those targeting the constant region of immunoglobulin G (IgG), called rheumatoid factors (RF). Despite this link, RFs can also be found in other disorders and the healthy population, which hampers its use as a diagnostic tool. We recently showed that a subset of RA-derived RFs target a distinct epitope on the IgG-Fc, a feature that is currently not used in the clinic.
Methods We determined immunoglobulin M (IgM)-RF levels specific against an RA-associated epitope (using our engineered next-generation RF antigen ‘T3-17’) in a prospective cohort of 475 patients with seropositive (for IgM-RF or aCCP) arthralgia that were followed for 5 years or until the development of arthritis.
Results The presence of RFs targeting T3-17 was more strongly associated with progression to arthritis in comparison to traditional RF measurements. Within the group of patients positive for T3-17 RF the risk of arthritis development was increased as compared with wild-type RF, HR=3.2 (95% CI 2.4 to 4.3) vs HR=2.2 (95% CI 1.7 to 3.0). Predictive power of T3-17 RF was improved in combination with aCCP titres, HR=6.4 (4.7–8.7) vs HR=5.1 (3.9–6.8). This combination performed better than aCCP detection on its own.
Conclusion The detection of disease-specific RF is feasible and seems to improve the diagnostic power of RF and should be considered to be implemented in the clinic.
- Rheumatoid Factor
- Rheumatoid Arthritis
- Sjogren's Syndrome
- Autoantibodies
Data availability statement
Data will be made available upon reasonable request
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Common rheumatoid factor (RF) testing is greatly limited by its regular occurrence in conditions other than rheumatoid arthritis (RA). We recently discovered that the epitopes that are targeted by RFs are disease specific.
WHAT THIS STUDY ADDS
Here we show that testing for RA-specific RFs greatly improves the predictive power of this test for the development of arthritis in a cohort of seropositive arthralgia.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Our study shows great promise for the improvement of RF testing through the detection of disease-specific RFs.
Introduction
Rheumatoid arthritis (RA) is an autoimmune inflammatory disorder that primarily affects the joints and bones and that is hallmarked by the presence of several distinct classes of autoantibodies. These are, in particular, antibodies against citrullinated proteins (anti-CCP) and antibodies against the constant region of immunoglobulin G (IgG), which are called rheumatoid factors (RF). These autoantibodies can already and often do emerge in the preclinical stage and have been found to be present in patients up to 10 years before established disease.1 Both anti-CCP and RF are part of the American College of Rheumatology/European League Against Rheumatism 2010 classification criteria for RA2 and are routinely tested on suspected disease. Whereas they share similar sensitivity, testing for anti-CCP is more specific than for RF.3 This is in part due to the presence of RF not limited to RA,4 but also because the RF response is highly heterogeneous.
Conventional RF testing is usually performed with agglutination assays or detection of immunoglobulin M (IgM)-RF bound to immobilised IgG molecules. Regardless, RF can occur as any given isotype, even though it is most commonly found and tested as IgM.5 Next, RFs are known to bind to multiple epitopes on the IgG-Fc and some of these epitopes are associated with specific disorders,6 7 a feature that is currently not used in the clinic. With this in mind, detection of RA-specific RFs may improve the specificity of these autoantibodies for predictive purposes.
We have shown previously in an explorative study that an epitope within the CH3 domain of IgG, captured by our IgG-Fc target ‘T3-17’, was strongly associated with RA.6 Preliminary results showed that detection of RF specific for this epitope in patients at risk for arthritis was linked with the presence of aCCP and showed a trend towards arthritis development. Here we validate these findings in a larger cohort of patients with seropositive arthralgia and show that detection of RA-specific RFs is feasible and improves the predictive value of RF testing.
Patients and methods
Study population
Baseline samples from the Reade seropositive arthralgia cohort (described previously8 9) were used. Patients who suffered from arthralgia and presented with a positive aCCP and/or RF test were enrolled from 2004 onwards at the Reade rheumatology clinic in Amsterdam. Enrolled patients were physically examined by a trained physician and a senior rheumatologist at the first study visit to ensure absence of arthritis. Furthermore, patients were checked for the absence of hand and feet erosions by X-ray. Patients were followed for 5 years after inclusion with yearly check-ups, or sooner on suspected arthritis. Arthritis was diagnosed if a patient presented with at least one swollen joint (out of 44 examined) on examination by a trained physician.
Determination of RF and aCCP levels
RF reactivity was determined in ELISAs as described previously.6 Briefly, IgG wild type (WT) or IgG T3-17 was directly coated overnight. Serum-derived IgM-RF was allowed to bind and detected with mouse a-IgM-HRP (MH-15; Sanquin). RF levels were quantified based on a pooled reference serum (RELARES) which has arbitrarily been set to contain 200 AU/mL of IgM-RF. RF levels against both targets were also determined for 224 healthy donors (online supplemental figure 1A). Levels of approximately 14 AU/mL against WT and 3 AU/mL against T3-17 conferred 95% specificity within this group (95% CI 91% to 97%). These estimates are similar if slightly lower than the optimal cut-offs determined by the receiver operating characteristic (ROC) analyses (14 vs 19 and 3 vs 3.8 AU/mL, respectively). aCCP levels were determined by an in-house ELISA using coated citrullinated peptides or by commercially available ELISA according to the manufacturer’s protocol (Axis Shield, Dundee, UK) as described previously.1 8 All aCCP levels were normalised to the latter test.
Supplemental material
Statistical analysis and visualisation
Data were analysed by logistic regression and ROC analysis using R V.4.0.3. The continuous input variables were log transformed, RF and/or aCCP levels and development of arthritis within 4-year follow-up were used as categorical response. Samples lacking 4-year follow-up were excluded from the analysis, resulting in 392 patients (table 1). Graphs were made and survival analyses were performed using GraphPad Prism V.9.
Baseline patient characteristics of the seropositive arthralgia cohort
Results
Patient characteristics
We determined RF reactivities for 475 patients with seropositive arthralgia who had at least 2-year follow-up after inclusion (table 1). For subsequent regression analyses, only the patients with 4-year follow-up were included (n=392). Patient characteristics can be found in table 1. Of these patients, 167 (42.6%) developed arthritis within the duration of the study after a median time of 11.5 months (IQR=17.4) after inclusion.
Baseline patient characteristics did not differ between the two groups (arthritis vs no arthritis; table 2). However, patients who later progressed to arthritis showed significantly higher frequencies of IgM-RF (p<0.01) and aCCP (p<0.05) positivity and had significantly higher C reactive protein levels (p<0.05).
Baseline characteristics of the seropositive arthralgia cohort split according to arthritis development during follow-up
IgM-RF levels against the T3-17 epitope are increased in the group that develops arthritis within 4 years
We previously defined an epitope, captured by IgG target T3-17, that was strongly associated with RA, but not primary Sjögren’s syndrome (SS) or healthy donors.6 In order to evaluate the potential of this new target for identifying patients at risk for development of RA within the above described cohort, we determined both WT and T3-17 IgM-RF levels in these patients. IgM-RF (WT) levels are higher in the group that developed arthritis within 4 years after inclusion (figure 1A,B). This difference between the groups is larger, when tested for reactivity against target T3-17, that is, the median levels are 2.5-fold and fourfold higher in the arthritis group versus non-arthritis group for WT and T3-17, respectively.
The development of arthritis in patients with seropositive arthralgia is associated with T3-17 RF. (A) Schematic representation of and IgM-RF levels against IgG1 WT and T3-17. For the latter, 16 human to mouse amino acid replacements were made in the IgG1-Fc region, which were determined to disrupt most RF binding except for those targeting the upper CH3 in the area indicated. IgM-RF levels were determined for all patients with at least 4-year follow-up after inclusion (n=392), after which they were grouped according to progression to arthritis within that time. (B) Correlation between WT RF and T3-17 RF levels for patients who did and did not develop arthritis within 4-year follow-up (n=225 and 167). (C) AUC values determined by linear regression analysis of arthritis development in relation with WT RF, T3-17 RF or aCCP, or a combination. (D) Arthritis-free survival in RF or aCCP-positive and negative patients (n=475). For the last graph, the positive group had either aCCP or T3-17 RF or both. Discriminative cut-offs for WT RF, T3-17 RF and aCCP were determined based on linear regression analysis (19.05, 3.8 and 15.4 AU/mL, respectively). aCCP, anti-citrullinated protein antibodies; AUC, area under the curve; IgG, immunoglobulin G; IgM, immunoglobulin M; RF, rheumatoid factor; WT, wild type.
Combination of RF against T3-17 and aCCP improves prediction of arthritis
To study the relation between RF reactivity against the RA-specific epitope T3-17 and the development of arthritis we performed logistic regression analysis (figure 1C). Presence of WT RF was significantly associated with progression to arthritis within 4-year follow-up (p<0.001). ROC analysis yielded an area under curve (AUC) of 0.59. This is in line with the results obtained previously within the same cohort for 2-year follow-up.7 RF against T3-17 is also significantly associated with arthritis development (p<0.001), but yields a higher AUC of 0.71. This indicates that RF against T3-17 shows improved predictive power for arthritis over WT RF.
Antibodies against CCP have previously been shown to correlate strongly with the progression to arthritis,9 10 so we therefore also included aCCP in our analysis. In line with previous results, aCCP is significantly associated with arthritis (p<0.001) and corresponds to an AUC of 0.72. This is only slightly improved on by combination of aCCP and WT RF (AUC of 0.73), whereas the combination of aCCP and T3-17 RF yields the highest AUC of 0.76, suggesting an improvement of using aCCP in combination with our new RF test over previous methods as well as only measuring aCCP.
Positivity for RF T3-17 is associated with an increased risk of arthritis
Based on the logistic regression analysis we determined the most discriminative cut-off values for WT RF, T3-17 RF and aCCP, which were 19.05, 3.8 and 15.4 AU/mL, respectively. Using these cut-off values resulted in positivity in the non-arthritis versus arthritis group of 41.3% vs 67.6% for WT RF, 27.5% vs 63.5% for T3-17 RF and 49.8% vs 85.6% for aCCP, respectively (online supplemental figure 1B). Applying these same cut-off values for data we previously obtained for patients with primary SS, 80% of patients with SS were positive, whereas for T3-17 RF this value was much lower, namely 27.2% (online supplemental figure 1A). Using these cut-offs, we also performed survival analyses. As reflected in the arthritis-free survival curves (figure 1D), positivity for WT RF yielded an HR of 2.2 (95% CI 1.7 to 2.9) over patients negative for WT RF, whereas patients positive for RF against T3-17 yielded a higher HR of 3.2 (95% CI 2.4 to 4.3), demonstrating the enhanced predictive power of T3-17. Patients positive for aCCP had an HR of 5.1 (95% CI 3.9 to 6.8) relative to those negative for aCCP. Predictive power could be improved on by combining aCCP and T3-17, which yielded an HR of 6.4 (95% CI 4.7 to 8.7) for the group positive for either versus the group negative for both. The survival curves made using the clinical RF test, with the manufacturer’s cut-off, showed very similar curves for both the RF positive and negative groups, likely owing to the low specificity and high cut-off of this test (online supplemental figure 1C). Taken together, these results show that detection of RFs binding to IgG1 T3-17 is related to arthritis development in at-risk patients.
Discussion
Screening for autoantibodies is one of the main serological, diagnostic tools to confirm suspected RA. Even though RF testing is performed routinely and remains an important part of the diagnostic criteria of RA, it lacks specificity as compared with the other main class of autoantibodies, aCCP. The aim of this study was to test if our recently described next-generation RF assay6 would perform better in predicting arthralgia in an at-risk population than conventional RF assays. We showed that T3-17-specific RFs were more tightly associated with the development of arthritis within 4 years in a cohort of patients with seropositive arthralgia compared with traditional WT RF measurements.
The presence of aCCP and RF are given the same weight in the classification criteria for RA,2 despite the lower specificity of RF. This may be improved by testing for T3-17-specific RFs instead, as we found these possess enhanced predictive power over WT RF, and approaching that of aCCP. Combined testing for aCCP and RF even enhanced the value of testing for aCCP alone, which corroborates earlier findings in this cohort.8 Furthermore, a more specific RF test would be immensely valuable for patients with suspected RA that test negative for aCCP, but positive for RF. As demonstrated recently,6 RFs in healthy individuals have a pattern of specificities that differs from that of patients with RA, and is presumably also not predictive of progression towards RA. Measurement of RFs that are more tightly associated to RA using T3-17 therefore is expected to reduce false-positive results.
A major limitation of this study is its inclusion requirements. Patients were selected based on seropositivity for aCCP and/or RF, and for that reason T3-17 may perform differently in a population where no such biases were introduced. The performance of T3-17 RF testing should be further assessed in other populations, preferably it should be tested next to routine screening for RF. Regardless, our results show great promise for the improvement of RF testing by only detecting RFs binding to RA-specific epitopes.
Data availability statement
Data will be made available upon reasonable request
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the Ethics Committee of Slotervaart Hospital and Reade (Amsterdam, the Netherlands). Participants gave informed consent to participate in the study before taking part.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors Study design: NO, TR. Data acquisition and analysis: NO, PO-dH, DK, TR. Writing of the manuscript: NO, TR. Revision of the manuscript: LvB, DvS, GW. Guarantor: TR.
Funding This work was funded by the Dutch Arthritis Foundation (grant 17-2-404).
Competing interests DvS, GW and TR are inventors on a patent application based on the use of bioengineered IgG targets for the characterisation of rheumatoid factor reactivity patterns.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.