Background Adverse drug reactions (ADR) associated with metothrexate (MTX) use in rheumatoid arthritis (RA) patients frequently lead to different complications and MTX withdrawal [1] especially without folate supplementation.

Objectives To reveal genetic markers associated with MTX adverse reactions in RA patients and to estimate their prognostic value.

Methods In the observational trial of RA patients (n=500: 405 women, 95 men) treated with disease modifying antirheumatic drug MTX was used in 30.6% (153/500) of patients. MTX withdrawal due to the adverse events occurred in 41.2% (63/153; CI₉₅ 33.7-49.1%) of them. This level of ADR frequency associated with WTX use was considered as MTX adverse reactions pretest probability (P_pre). Association between MTX ADR and a number of the genetic markers (AB0, Rh₀, MN, P₁blood groups; haptoglobin; HLA-A, -B, -C, -DR, -DQ locuses, supertypes HLA-Bw4, HLA-Bw6, HLA-DR51-53; sensitivity to phenylthiocarbamide; dermatoglyphic characteristics) was investigated. All markers were dichotomic (marker⁺, marker^-). Statistical significance of the revealed association was estimated by Fisher exact test. Likelihood ratio of positive (LR⁺) and negative (LR^-) tests and prognostic odds ratio (pOR) as well as post-test probability (P_post) of MTX adverse reactions were calculated.

Results Gastrointestinal ADR and hepatotoxicity were registered in 13.1% (20/153) of patients, skin and mucous ADR in 12.4%, haematological abnormalities in 10.5% and infections in 9.8% of patients. MTX adverse reactions were significantly more common in the case of following phenotypes:

• – P₁^– vs P₁⁺ 75.0% (9/12) and 12.5% (1/8), p_2-t=0.0198;

• – HLA-A10⁺ vs HLA-A10^- 58.6% (17/29) and 32.1% (25/78), p_2-t=0.0152;

• – HLA-C2^– vs HLA-C2⁺ 46.7% (28/60) and 0.0% (0/7), p_2-t=0.0359 with adjustment of zero frequency by J. Haldane;

• – HLA-DR3⁺ vs HLA-DR3^–68.8% (11/16) and 33.3% (15/45), p_2-t=0.0194;

• – HLA-DQ2⁺ vs HLA-DQ2^– 61.9% (13/21) and 32.5% (13/40), p_2-t=0.0333.

On the basis of these findings operational parameters of the revealed markers were determined as predictors of A₁ outcome (MTX ADR⁺) and A₂ outcome (MTX ADR^-):

• – P₁: pOR=21.0; P₁^–: LR⁺=3.0, P_post=61.7%; P₁⁺: LR^–=0.14 P_post= 8.9%;

• – HLA-A10: pOR=3.0; A10⁺: LR⁺=2.2, P_post=60.7%; A10^–: LR^-=0.73, P_post= 33.8%;

• – HLA-C2: pOR=13.1; C2^–: LR⁺=1.2, P_post=45.7%; C2⁺: LR^–=0.09, P_post= 5.9%;

• – HLA-DR3: pOR=4.4; DR3⁺: LR⁺=3.0, P_post=75.5%; DR3^–: LR^–=0.67, P_post= 31.9%;

• – HLA-DQ2: pOR=3.0; DQ2⁺: LR⁺=2.0, P_post=58.4%; DQ2^–: LR^-=0.67, P_post= 31.9%.

There was no association between sensitivity to phenylthiocarbamide or dermatoglyphic characteristics and MTX adverse reactions.

Conclusions Revealed phenotypes are useful as supplementary predictors of MTX adverse reactions in RA patients in the case of uncertain prognosis on the basis of clinical variables [2]. Availability of several independent predictors leads to considerable increase of the post-test probability of prediction of ADR, associated with MTX, oversteping the significant threshold levels of prediction: P_post≥95% for the approval of the hypothesis (A₁ outcome) or P_post≤5% for the acceptance of the alternative hypothesis (A₂ outcome).

1. Bologna C, et al. Long-term follow-up of 453 rheumatoid arthritis patients treated with methotrexate: an open retrospective, observational study. Br J Rheum 1997;36:535-40.

2. Hoekstra M, et al. Factors associated with toxicity, final dose, and efficacy of methotrexate in patients with rheumatoid arthritis. Ann Rheum Dis 2003;62:423-426.

Disclosure of Interest None Declared