Background/Purpose:

About 30-40% of patients with RA do not respond to MTX, the first-line therapy in RA. Early identification of responders may allow the use of alternative DMARDs in patients unlikely to respond, thus preventing long-term damage. CD39, an ectonucleotidase highly expressed on regulatory T cells (Tregs), is responsible for the production of adenosine, an important anti-inflammatory mediator of MTX action. CD39 expression on Tregs improves their suppressive capacity. Therefore, we aimed to study the role of CD39⁺ regulatory T cell (CD39⁺ Treg) frequency as a biomarker for MTX responsiveness in RA.

Methods:

Patients with active RA (fulfilling 2010 ACR/EULAR classification criteria and having DAS28-CRP >3.2) who were naive to DMARDS were enrolled. Frequency of CD39⁺ Tregs (CD39⁺CD4⁺CD25⁺FoxP3⁺ T cells) and CD39⁺CD4⁺CD25⁺ cells was determined by flow cytometry in peripheral blood before start of therapy. After 4 months of treatment with MTX monotherapy (no corticosteroids were given), patients were classified into two groups based on EULAR response criteria: responders (moderate/good response; MTX-R) and non-responders (MTX-NR). All patients who needed rescue therapy after 2 months were classified as non-responders. All patients were genotyped for single nucleotide polymorphism (SNP) rs11188513 in CD39 gene using TaqMan probe method. The data (median and IQR) was analyzed using non-parametric tests.

Results:

Among the 70 patients who completed at least 2 months follow-up (60 females, median age: 39 years and median disease duration: 24 months), 53 patients were RF positive and 57 were anti-CCP positive. The baseline median DAS-28 CRP was 5.15 (4.45-5.89). After 4 months of follow-up, 54 patients were classified as MTX-R and 16 as MTX-NR.

There was no difference in the two groups as regards age, disease duration, disease activity or baseline frequency of Tregs [1.4% (1.0%-2.3%) vs 1.1% (0.7%-1.8%), p = NS]. However, the baseline CD39⁺Treg frequency was significantly higher in the MTX-R compared with the MTX-NR [78.0% (68.3%-87.4%) vs 67.8% (24.7%-84.3%), p < 0.05]. At a cut-off of CD39⁺ Treg frequency of 75.1%, the positive predictive value (PPV) to identify responders was 86%.

As Foxp3 staining takes time we also saw if the frequency of CD39⁺CD4⁺CD25⁺ T cells could be utilized for prediction of MTX response. The baseline CD39⁺CD4⁺CD25⁺ T cells frequency was also significantly higher in the MTX-R compared with the MTX-NR [43.6% (36.0%-52.7%) vs 32.6% (25.0%-39.7%), p < 0.01]. At a cut-off frequency of 41.7%, PPV to identify responders was 91%. MTX treatment did not alter the CD39⁺ Treg frequency [75.9% (61.8%-87.0%) vs 67.2% (51.3%-84.0%), p = NS].

CC genotype at SNP rs11188513 of CD39 gene was associated with a significantly lower frequency of both CD39⁺ Tregs (p < 0.01) as well as CD39⁺CD4⁺CD25⁺ T cells (p < 0.01).

Conclusion:

Higher frequency of CD39⁺ Tregs in the peripheral blood is associated with response to MTX in RA and hence, this should be considered as a potential biomarker for prediction of response to MTX treatment.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/cd39-positive-regulatory-t-cells-as-a-biomarker-of-responsiveness-to-methotrexate-in-rheumatoid-arthritis/