Serum Amyloid A: An Emerging Biomarker in Early RA?

Serum amyloid A levels may be more accurate than traditional measures of disease activity in RA, including C-reactive protein levels and erythrocyte sedimentation rate.

A single measure for accurately measuring disease activity of rheumatoid arthritis (RA) has remained elusive, and rheumatologists have relied upon composite disease activity scores, which have limitations in guiding clinical decision-making. Serum amyloid A (SAA), a systemic acute-phase reactant, has shown some promise as a biomarker for RA, but data were lacking. A new study that examined SAA among large groups of RA patients enrolled in the Treatment of Early RA (TEAR) study found that it may be a better biomarker of RA disease activity than C-reactive protein (CRP), especially during treatment with tumor necrosis factor (TNF) antagonists.

“The advantages of SAA as a biomarker of disease activity include the rapid production and exceptionally wide dynamic range of the SAA response, while CRP remains normal,” said the study’s lead investigator Yong Gil Hwang, MD, Division of Rheumatology and Clinical Immunology, University of Pittsburgh, Pittsburgh, Pennsylvania, in an email interview with Rheumatology Advisor. “Therefore, we investigated whether SAA levels accurately reflect RA disease activity and whether there are differences in reductions of SAA levels with various RA treatments.”

In the study, 755 patients with early RA (< 3 years’ disease duration, < 2 months of prior disease-modifying anti-rheumatic drug [DMARD therapy], active disease with ≥ 4 swollen joints and 4 tender joints using a 28-joint count) had their serum samples analyzed at initiation of treatment and at follow-up visits at 24, 48, and 102 weeks. Patients were randomly assigned to one of 4 treatment groups: immediate treatment with methotrexate (MTX) plus etanercept (ETN); immediate oral triple therapy (MTX plus sulfasalazine [SSZ] plus hydroxychloroquine [HCQ]); or step up from MTX monotherapy to either MTX plus ETN or MTX plus SSZ plus HCQ at week 24 (start of the step-up period) if their disease activity score of 28 joints based on erythrocyte sedimentation rate (DAS28-ESR) was ≥ 3.2.

Significant differences were observed with SAA levels by visit and treatment arm. RA patients who were treated with ETN/MTX had greater reduction in SAA than those treated with oral DMARDs, even when accounting for disease activity. SAA levels were decreased by an average of 66 ranks as a result of ETN/MTX treatment versus triple oral therapy. Similar results were observed with serum CRP by visit and by treatment arm, but with an even greater reduction of 144 ranks following ETN/MTX treatment versus triple oral therapy. Additionally, models of DAS28-ESR using SAA were shown to be better than those using CRP across all patient groups and time periods.

“SAA levels modeled RA disease activity better than CRP levels and it may be a better biomarker for RA disease activity than CRP levels, especially during treatment with an anti-TNF inhibitor such as etanercept,” said Dr Hwang.

As a biomarker for early RA, SAA has many desirable traits, such as rapid production (1000-fold increase during acute inflammation), wide dynamic range, much shorter half-life than CRP, and its direct involvement in synovial inflammation and joint destruction. SAA may also be less susceptible to the varying mechanisms of action of different treatments.

Summary & Clinical Applicability

“It is well known to practicing rheumatologists that RA disease activity cannot be measured accurately by a single measure and current composite disease activity scores are not perfect to guide treatment decisions yet,” said Dr Hwang. “Our study suggests that SAA concentrations can be an integral part of the clinicians’ repertoire to assess RA disease activity and to monitor therapeutic response.”

Limitations & Disclosures

Dr Hwang and colleagues reported several limitations, including the lack of a gold standard to objectively measure RA disease activity and the influence of non-RA factors on biomarkers currently used to assess RA activity. Additionally, patients in the study all had early active RA, limiting the applicability of the findings to other RA patients, especially those with long disease duration or mild disease activity.

Nevertheless, the findings were significant enough that the authors are interested in conducting additional research to determine the practical usefulness of SAA as a biomarker for measuring RA disease activity in routine practice settings.

No conflicts of interest were reported.


Hwang YG, Balasubramani GK, Metes ID, et al. Differential response of serum amyloid A to different therapies in early rheumatoid arthritis and its potential value as a disease activity biomarker. Arthritis Research & Therapy. 2016;18:108.

Related Articles